Version 1.0 · Public Research Edition · TLP:CLEAR · Brandefense Country Report · Germany 2026
About this editionPrepared with AI assistance. Evidence and limitations are stated beside the findings.
Chapter 1
Executive summary
This chapter answers
Which findings justify immediate decisions for German organisations?
In short
Germany's evidence supports three priorities: identity security, recovery and supplier oversight. Ransompedia records 457 German listings, exceeding the full-year 2025 total of 311. Public disclosures distinguish data theft, credential misuse and service disruption. Recent university incidents, industrial-cooling warnings and provider notices reinforce that distinction. The exposure snapshot supports account validation across the selected portfolio. Confirm your own account and supplier alerts, rehearse recovery, and assign separate owners to operational continuity and notification decisions.
This report combines Brandefense's retained underground material and ransomware listings with public reporting. Posts, listings, public incidents and domain exposure observations are separate populations. They are never added into a national attack total. [S1][S2]
The requested window ends on 22 November. Evidence reflects what was available when reviewed on that date; the cut-off day was still in progress during collection. The closing fortnight is treated separately in §5.7. Publication dates do not establish when an intrusion began.
1. Extortion visibility is distributed across many claimant labels. Germany has 457 listings against 311 in all of 2025. Qilin, SafePay and The Gentlemen account for 161, or 35.2%, of the current-window listings. The remaining 296 belong to other labels. This compares a partial year with a full year; it is not a matched-period growth rate. A listing establishes a public criminal assertion, not encryption, theft or financial loss. [S2]
2. Confirmed identity misuse deserves a different response from an advertisement. TUM's disclosure describes central-login credentials, including an administrator's credential, being used against a campus system. Its recovery update required password changes and two-factor authentication. That supports account and session containment for an organisation with comparable evidence. It does not establish that every university account was stolen. [S11][S12]
3. Supplier events can continue as fraud after technical restoration. An official hospital-related warning describes direct extortion messages to patients using stolen information. A separate hotel-software warning describes follow-up fraudulent messages. These are documented routes from information loss to customer harm. Legal, customer operations and fraud teams should join the response before a system returns. [S16][S18]
4. Availability, data loss and physical-process safety require separate evidence. Deutsche Bahn confirmed February disruption of booking and information services. November's official cooling warning concerns industrial controls. The two justify different continuity exercises; neither supports assigning a common actor or assuming that a website outage means operational-system access. [S7][S8][S21]
5. Current credential offers require owned-account validation. Of 249 observed domains, 107 had a credential offered for sale. The selected portfolio records at least 636,397 device associations and 125,422,358 credential-record associations. These overlapping observations do not count unique people or confirmed intrusions. One selected domain has no complete exposure record. [S87]
Four decisions for leaders
- Validate identity alerts against your own environment. Establish an alert route, confirm account ownership and freshness, inspect privileges, and revoke sessions where warranted. Record a disposition for each alert (§3.4, §12).
- Test restoration and minimum service together. Restore a critical service in isolation and demonstrate a workable fallback while the normal identity or booking system is unavailable (§5–6).
- Bring suppliers into incident decisions early. Identify who holds customer records, who controls remote maintenance, and who can provide timely evidence. Rehearse notification with a critical supplier (§8, §10).
- Separate communications by the facts established. Say whether a claim, an outage, data access or operational manipulation is confirmed. Give an update time and an accountable spokesperson (§6, §12).
The year in numbers
Observed Germany-associated posts in the requested window
What it does not measureA national incident or unique-victim count
German ransomware listings dated within the requested 2026 window
What it does not measureConfirmed intrusions or unique people
German listings for all of 2025
What it does not measureA matched-period baseline
Listings under Qilin / SafePay / The Gentlemen labels
What it does not measureUnique operators or their success rates
Listings carrying the manufacturing sector label
What it does not measureA national manufacturing incident rate
Listings for France / Netherlands / Poland / Austria
What it does not measurePopulation-adjusted risk or detection effectiveness
Selected domains with exposure observations
What it does not measureIndependent legal entities or national prevalence
Observed domains with a current credential offer
What it does not measureAccounts sold or confirmed intrusions
Comparison countries were selected as examples to provide context only; they do not imply a ranking or political assessment.
Sources: Odin by Brandefense [S1], Ransompedia [S2] and the exposure snapshot [S87]. Country labels and group names describe the listing population. Industry labels do not establish the precise operating activity of every listed organisation.
Key takeaways
- 01
Use extortion listings to prepare recovery, not to count confirmed intrusions.
- 02
Validate identity evidence before choosing disruptive containment.
- 03
Supplier, customer and operational responses need owners alongside technical recovery.
Report map
Twelve chapters, five reader paths. Follow the question that matches your responsibility; evidence limits travel with every conclusion.
The evidence labels are observation, claim, official confirmation and analyst inference. An observed advertisement remains a claim about its contents. A confirmed incident can leave the actor, data quantity and entry route unresolved.
Contents and reading routes
| Chapter | Question |
|---|---|
| 1 Executive summary | Which decisions should come first? |
| 2 Context | Which German operating conditions shape the response? |
| 3 Organisation exposure | Which measured signals need validation? |
| 4 Underground activity | What do the retained posts actually establish? |
| 5 Incidents by sector | What did organisations and authorities confirm? |
| 6 DDoS and operational technology | How do availability claims differ from operational impact? |
| 7 Actors and relationships | Which associations are supported, and by which source? |
| 8 Fraud economy | How can stolen information become a second incident? |
| 9 Banking | How do financial-domain observations inform response? |
| 10 Regulation | Which reporting clocks and responsibilities apply? |
| 11 Outlook | Which conditional scenarios justify preparation? |
| 12 Action plan | Who acts, and what demonstrates completion? |
| Appendices A–D | Scope, evidence, sources and definitions |
Chapter 2
Germany's digital and threat context
This chapter answers
Which operating conditions shape a proportionate German response?
In short
German organisations face overlapping identity, supplier and availability risks. National authorities describe these through different reporting systems and observation windows. Their statistics cannot be added to underground listings. Responsibility also spans information security, data protection and sector supervision. Map the service, the legal entity, the supplier and the competent reporting channel before an incident starts.
2.1 Separate the clocks before comparing the figures
The BSI's 2026 annual assessment covers 1 July 2025–30 June 2026. Its publication in November therefore supplies recent analysis of an earlier observation window. It is not a census through this report's cut-off. [S3]
The BKA's May publication reports 1,041 ransomware complaints for 2025. Ransompedia's 457 listings describe a different year, source and unit. Neither is the missing portion of the other. A complaint can remain unpublished, while a listing may never become a confirmed incident. [S4][S2]
The Berlin data-protection authority recorded 1,550 breach notifications during January–September 2026, split between 1,303 non-public and 247 public-sector notifications. These include personal-data incidents beyond malicious intrusions. They are regional notifications, not a national cyberattack count. [S35]
The following distinctions prevent false comparisons.
| Population | Unit | Appropriate decision |
|---|---|---|
| Underground posts | Published messages and advertisements | Validate claims relevant to your assets |
| Extortion listings | Public criminal assertions | Test recovery and disclosure readiness |
| Police complaints | Offences reported under police definitions | Understand reported criminal activity |
| Data-protection notifications | Reported personal-data breaches | Assess risk to affected people |
| Confirmed disclosures | Organisation or authority statements | Compare mechanisms and response choices |
2.2 Shared services create shared response dependencies
The public cases show several forms of dependency: campus identity, external customer-data services, communications providers and remotely maintained equipment. A single technology relationship can involve several legal entities and different operational owners. Count affected customers separately from incidents at the provider.
Our interpretation is practical. A service inventory should show the business function, identity dependency, data custodian, supplier contact and fallback mode. A contract list alone does not show how an outage reaches a ward, a passenger or a laboratory.
For a university, the necessary question is whether central authentication can be restored safely before registration deadlines. For a manufacturer, it is whether supplier access can be suspended while production remains safe. For a public body, it is whether citizens can obtain essential services while document review continues.
These are planning questions derived from the cases. The report does not claim that every organisation in a sector shares the same architecture or has failed to implement these controls.
2.3 Politically framed activity requires careful attribution
BSI and BfV describe messenger phishing against high-profile political, military, diplomatic and journalistic users. BSI considers state support plausible; its retrospective identifies other countries' attribution separately. This report reports each assessment as the source's position. [S5][S6]
A political target does not establish a political motive. In an October party-account case, the investigating authorities reported no indication of a political background at that stage. The appropriate lesson is to secure externally hosted mail and recovery routes across organisations, regardless of affiliation. [S41]
The report takes no position on elections, political parties, government decisions, Russia–Ukraine, Israel–Gaza or China. Language, a claimed cause and an actor's self-description are not evidence of nationality or state direction. No election outcome is attributed to a cyber incident here.
2.4 Regulation is already an operating requirement
Germany's NIS2 implementation took effect on 6 December 2025. Applicable organisations should therefore establish scope and operational obligations against the law in force. They should not carry over a neighbouring country's legislative timetable. [S30]
Data-protection reporting and financial-sector reporting have their own triggers. A supplier notice may start a controller's awareness assessment, while a service outage may require a separate sector determination. Chapter 10 distinguishes those decisions and identifies completion evidence.
2.5 What this context changes for leaders
Use the same incident record for shared facts, while recording each reporting decision separately. The record should distinguish the time of awareness, the service affected, the people potentially at risk and what remains unverified.
Assign a business owner to each critical dependency. Security can establish whether a credential was used, but the business owner must decide which minimum service remains acceptable. Legal and the data-protection officer determine the applicable reporting route and rationale.
Key takeaways
- 01
National statistics and underground listings describe different populations and periods.
- 02
Shared technology can create several affected organisations without several independent intrusions.
- 03
Establish legal scope and reporting ownership before an event forces the decision.
Chapter 3
National exposure: a selected domain portfolio
This chapter answers
Which measured exposure signals require validation by the selected organisations?
In short
107 of 249 observed domains had a credential listed for sale when checked. The selected portfolio records at least 636,397 device associations and 125,422,358 credential-record associations. These are overlapping observations, not unique people or confirmed intrusions. One of the 250 selected domains has no complete exposure record. Validate current listings, investigate relevant employee-device associations and route other account evidence to the appropriate service owner.
3.1 Universe card
| Item | Definition |
|---|---|
| Selected population | 250 purposively selected domains across 10 sector groups; 25 candidates per group |
| Observed population | 249 domains; industry and technology has 24 observations from 25 selected domains |
| Selection | Service and sector relevance; no revenue ranking or nationally representative sample |
| Unit | Domain association; legal ownership and independence of the represented organisations remain unvalidated |
| Snapshot | 21–22 November 2026; cumulative device, credential, document and market observations, not a 2026-only incident population |
| Lookalike period | Previous 30 days; similar naming does not establish abusive use |
| Mention observations | Previous 365 days ending 18 November 2026; separate from the exposure snapshot |
| Overlap | Records can recur across sources, domains and people; related services can share an operator |
| Passive results | No completed hygiene dataset; missing security-control measurements remain unavailable |
The portfolio is selected, not representative. Its domain totals do not establish a count of independent legal entities. A missing record remains outside the observed denominator. The exposure snapshot and the report's incident window describe different periods. [S87]
3.2 Current listings and measured signals
The national Atlas headline is 61 out of 100 for the 249 domains with complete exposure observations. This is a portfolio figure, not national prevalence. The table contains signal counts, with no sector or organisation ratings. [S87]
Cells show domains with the signal / observed domains in each sector. Current credential offers are shown in coral.
Source: exposure observations for 249 of 250 selected domains, 21-22 November 2026. Cumulative, overlapping domain associations; not independent organisations, unique people or confirmed intrusions [S87].
| Sector | Observed domains | Current credential listings | Listing history | Employee-device associations present | Document observations present |
|---|---|---|---|---|---|
| Public administration | 25 | 2 | 2 | 10 | 18 |
| Regional and local administration | 25 | 9 | 9 | 19 | 23 |
| Banking and finance | 25 | 13 | 12 | 20 | 21 |
| Insurance | 25 | 10 | 10 | 18 | 21 |
| Universities and research | 25 | 21 | 21 | 24 | 23 |
| Industry and technology | 24 | 11 | 11 | 22 | 23 |
| Energy and infrastructure | 25 | 4 | 4 | 16 | 14 |
| Transport and logistics | 25 | 9 | 9 | 11 | 11 |
| Consumer services and communications | 25 | 22 | 22 | 22 | 23 |
| Health and life sciences | 25 | 6 | 6 | 18 | 15 |
| ALL | 249 | 107 | 106 | 180 | 192 |
Source: Brandefense Blackhole associations and Atlas exposure observations, snapshot dated 21–22 November 2026 [S87]. Listing counts concern domains with an offer, not accounts sold or confirmed breaches.
3.3 Devices and credential records
- Employee-device associations
- Other device associations
Source: exposure observations for 249 of 250 selected domains, 21-22 November 2026. Cumulative, overlapping domain associations; not independent organisations, unique people or confirmed intrusions [S87].
| Sector | Observed domains | Device associations, at least | Employee-device associations | Credential-record associations |
|---|---|---|---|---|
| Public administration | 25 | 23,283 | 40 | 54,026 |
| Regional and local administration | 25 | 26,157 | 419 | 869,088 |
| Banking and finance | 25 | 80,163 | 698 | 669,977 |
| Insurance | 25 | 17,069 | 233 | 333,920 |
| Universities and research | 25 | 43,877 | 2,096 | 939,422 |
| Industry and technology | 24 | 90,446 | 4,854 | 3,281,078 |
| Energy and infrastructure | 25 | 10,580 | 266 | 158,849 |
| Transport and logistics | 25 | 83,168 | 1,583 | 681,446 |
| Consumer services and communications | 25 | 238,181 | 23,575 | 118,214,739 |
| Health and life sciences | 25 | 23,473 | 151 | 219,813 |
| ALL | 249 | 636,397 | 33,915 | 125,422,358 |
Credential records comprise 48,809,419 breach-database associations, 9,068,279 stealer-log lines and 67,544,660 combolist entries. They do not establish unique people, working passwords or recent access. [S87]
The device population contains 33,915 employee-classified associations and 602,482 other associations. Other associations are not verified customer devices. Shared service domains, historical records and repeated associations can contribute to the totals. Domain ownership must be validated before contacting an affected party. [S87]
Consumer services and communications account for 118,214,739 credential-record associations. Public-facing services can contain customer identities as well as staff accounts. This concentration does not establish equivalent intrusion exposure at every represented organisation. [S87]
3.4 Account validation and containment
Confirm that an account and service belong to your organisation. Check record age, account status and prior remediation. For active privileged accounts, inspect sign-ins, session grants and administrative changes. Revoke sessions and rotate affected credentials when the evidence warrants containment.
A password reset is not the only closure test. Sessions, added authentication methods and linked devices can require separate action. Customer or student accounts need their own response owner. Record whether the evidence was current, historical, unrelated or unresolved.
3.5 Documents, lookalikes and passive controls
The snapshot contains 5,921 document observations and 153,973 lookalike observations. Document association does not establish confidential content or unauthorised access. Similar domain names do not establish impersonation. Check the specific asset, ownership and abusive behaviour before escalation. [S87]
No completed passive hygiene dataset is available. The report publishes no sector hygiene index values or email-control prevalence. A missing result does not establish a missing control. Owners should validate legitimate mail senders, authentication alignment, certificates and service dependencies locally.
3.6 Response ownership
Keep current-account validation, employee endpoint investigation, customer protection and configuration review in separate response queues. Each queue needs an accountable owner, supporting evidence and a closure test. The portfolio's counts cannot determine which specific system requires disruptive containment.
Key takeaways
- 01
Current listings are account-validation leads, not confirmed intrusions.
- 02
Device and credential associations overlap and retain their domain-based denominator.
- 03
Assign response owners while preserving missing records and unvalidated ownership.
Chapter 4
Underground activity and claim interpretation
This chapter answers
What do the retained advertisements establish about access, documents and identity demand?
In short
The retained posts include advertised administrator access, organisational documents, project records and demand for German identity information. Six case cards show those propositions without adopting the sellers' claims. Two illustrated November offers belong to the recent-activity window. The observed post population is not a national incident census. No completed-sale estimate or crime-family ranking is established. Validate relevant claims against owned accounts and data, and distinguish an offer from possession, a repeat advertisement from corroboration, and demand from theft.
4.1 Universe card
| Item | Status |
|---|---|
| Requested window | 1 January–22 November 2026, by post time |
| Observed Germany-associated post population | 1,221,949 posts [S1]; no national incident count |
| Retained evidence | Forum posts discussing Germany and German-associated assets |
| Illustrated claims | 6 case cards, each paired with a redacted source quotation |
| Composition | Platform and language observations below; no national crime-family ranking established |
| Complete forum and market inventory | Unavailable |
| Ransomware listings | Separate population: 457 German listings [S2] |
The observations below establish what was published. They do not establish a complete underground population or verify the assets offered. A German reference can identify a claimed target, a customer group or a sales market. It does not by itself establish the organisation's location or the origin of every record. [S1]
The observed count and the source-content inventory answer different questions. Missing content does not establish absence of activity. Public disclosures and ransomware listings do not substitute for the post population. The six illustrated claims show different decisions; their number is not a prevalence estimate. [S1]
Monthly observed post population
- Observed Germany-associated posts
Source: Brandefense observed Germany-associated posts, 1 January-22 November 2026; cutoff day partial. Posts, retained claims and incidents are separate populations [S1].
| Month | Posts | Month | Posts |
|---|---|---|---|
| January | 429,473 | July | 24,813 |
| February | 417,736 | August | 41,766 |
| March | 118,617 | September | 33,190 |
| April | 43,761 | October | 31,879 |
| May | 37,662 | November through cut-off | 11,724 |
| June | 31,328 | Total | 1,221,949 |
Source: Odin by Brandefense [S1]. January and February account for 69.3% of the observed population. Posts can be advertisements, discussion, repeated offers or international material with a German association. A change in publication volume does not establish a change in successful attacks or national exposure. November remains partial.
The post count does not supply a unique-person denominator, a completed-sale estimate or a verified geographic location for every asset. Its value is visibility context. Use the source-checked claim cards and public disclosures for the specific response questions.
The overall post total and monthly counts agree. The platform breakdown remains seven posts short (§4.7). [S1]
4.2 Recent offers: a tenant account and organisational files
Case S
Advertised administrator access to a banking-related CRM tenant
- Claim
- A forum advertisement published on 8 November offers a full administrator account for a cloud CRM tenant associated in the text with a consumer bank. It describes customer contact information. [S1]
- What can be established
- The offer and the asserted privilege were present in the retained post. The material does not establish that the account worked, remained active or belonged to a German legal entity. A banking-related tenant claim is not evidence that the cloud platform's own infrastructure was accessed.
- Alternatives
- The account could be stale, misrepresented, limited to a supplier tenant or advertised without possession. No successful sale, customer population or bank-wide intrusion is established.
- Response
- The identity owner should validate tenant ownership, administrator assignments, recent sign-ins and session grants through authorised records. Containment should follow the validated account evidence, with customer-data review handled separately.
Case T
An offer of files from German organisations
- Claim
- A 16 November post offers data described as originating from German organisations and says proof is available. Other retained posts contain similar advertising during the same period. [S1]
- What can be established
- The source asserts possession and a German connection. The report does not verify the file origin, quantity, age or authenticity. Similar lists in later advertisements do not establish another theft or an independent confirming source.
- Alternatives
- Advertising can be copied; several bundles can contain overlapping or older records. A document naming a public body, supplier or bank does not establish access to that body's systems.
- Response
- Data owners should compare an authorised alert with their own document inventory and prior notifications. Record what is actually matched before assigning an incident date, affected population or reporting obligation.
- Held
- An advertisement asserts administrator access to a banking-related CRM tenant and describes customer information.
- Not established
- Usable access, a German legal-entity target, underlying platform intrusion, a completed sale or an affected-customer count.
- Held
- An advertisement offers files said to come from German organisations.
- Not established
- File authenticity, origin, freshness, a completed sale or a new intrusion each time similar advertising appears.
These November observations are publication events. They are not dated confirmations of two new intrusions. Neither is linked here to the university, political-party or bank disclosures in other chapters. The report has no concrete evidence of a shared operator or transaction chain. [S1]
4.3 Documents as a fraud resource
Case U
Project contacts and budget records offered for sale
- Claim
- On 2 February, a seller describes a database of regional project contacts and budget records from a German portal. [S1]
- What can be established
- The advertisement combines contact and project-finance categories. The claimed record count and domestic origin are not independently verified. No transfer from the portal or completed transaction is established.
- Alternatives
- Contact material may be older, public, copied or combined from several sources. A plausible field list does not establish a current vulnerability or complete customer database.
- Response
- Procurement and fraud teams should authenticate changes to project payments through a known channel. A relevant alert should be checked with the data custodian, including whether the same material was already disclosed.
Case V
A claimed release of administrative and financial files
- Claim
- A 6 October post announces a data release from an organisation it describes as located in Germany. It lists administrative and financial data, corporate email material and internal documents. [S1]
- What can be established
- The release assertion and document categories were published. The underlying files, asserted size, intrusion route and freshness are not verified here.
- Alternatives
- A release announcement can concern an old export, a supplier copy or repackaged material. It does not establish that the organisation's current systems remained accessible.
- Response
- The incident owner should establish document ownership and age before choosing containment. Fraud teams should monitor for plausible invoice, supplier and executive impersonation using those document categories.
- Held
- A seller describes project-contact and budget records associated with a German portal.
- Not established
- A verified record count, portal intrusion, current flaw, authentic files or a completed sale.
- Held
- A post asserts release of administrative, financial and corporate email material from an organisation described as German.
- Not established
- Authentic files, a new intrusion, continued access, verified volume or a completed transaction.
The analytical relationship is between useful context and a later deceptive request. Project budgets or internal correspondence can make a fraudulent message more credible. These posts do not establish that such a message was actually sent. The confirmed follow-on warnings in Chapter 5 provide separate evidence of that behaviour; they are not matches to these advertisements.
4.4 A vulnerability assertion and a demand signal
Case W
Automotive documents advertised with an access-control claim
- Claim
- A 17 February advertisement offers documents, including personal information, from a major automotive group described as German. It attributes acquisition to an insecure direct object reference, or IDOR, and separately offers the asserted exploit for sale. A second advertisement repeats the claim later that day. [S1]
- What can be established
- The seller makes a document and vulnerability claim. The post does not verify a live flaw, exploitation, file possession or a new automotive intrusion. The repeat advertisement is not another incident.
- Alternatives
- The route may be guessed, copied or inaccurately described; documents can originate from a contractor or historical export. The seller's terminology is not an independent technical assessment.
- Response
- An application owner receiving a relevant alert should review object-level authorisation and access logs in the affected owned service. Use an authorised test and preserve evidence of the actual result, rather than treating the advertisement as a vulnerability finding.
Case X
Demand for German identity and home-address information
- Claim
- A 5 January post asks for information about German people, emphasising physical home addresses and requesting other identity details if available. [S1]
- What can be established
- This is a request to obtain information. It does not establish that the author held a database, acquired records, completed a purchase or selected a particular victim.
- Alternatives
- The request may never have been fulfilled. It gives no supported number of people, source organisation or collection date.
- Response
- Treat a relevant identity alert as a data-protection and customer-safety question. Examine actual information disclosed and credible misuse, while avoiding a national affected-person count derived from a buyer's request.
- Held
- A seller advertises automotive documents and separately an IDOR exploit.
- Not established
- A real vulnerability, exploitation, possession of the documents, a sale or a second incident from the repeated advertisement.
- Held
- A buyer requests German identity information and emphasises real home addresses.
- Not established
- Possession, acquisition, a completed purchase, a specific affected person or an underlying intrusion.
The six examples distinguish advertised access, offered records, an asserted release, an asserted technical route and demand. These categories require different evidence. They should not be combined into a six-intrusion total. [S1]
4.5 Other retained claims and historical advertisements
The following additional observations widen the claim register without increasing the confirmed-incident population. Dates are publication dates. None validates possession, a completed transaction or the actor's asserted technical route. [S1]
| Post date | Role and assertion | Boundary and response |
|---|---|---|
| 1 February | Germany-labelled website database offered [S1] | The country label does not verify origin, age or an intrusion; the data owner should check any relevant alert. |
| 31 March | German-language forum-account dataset shared [S1] | Language is not residence; inconsistent scale descriptions are omitted. Validate account ownership and prior remediation. |
| 6 June | German business/contact dataset advertised as 2025 material [S1] | Older directory or public-source material is plausible; publication in 2026 does not establish a new intrusion. |
| 17 June | German consumer-contact dataset advertised as 2025 material [S1] | Historical trade claim, without a verified source organisation or breach date; assess actual data categories. |
| 12 August | Telephony service advertises a German-number capability restriction [S1] | Capability claim only, with no observed victim or successful impersonation; use independent call authentication. |
| 22 August | German mobile-number and birth-date dataset claimed [S1] | Origin, age and possession are unverified; do not assign a telecom-provider intrusion. |
| 7 September | A German retail/services group is named in a claimed customer-data event [S1] | The post calls it a 2026 event; no official confirmation or file authenticity is established. |
| 3 October | Financial-user records described as German offered for sale [S1] | Customer fields do not establish bank-system access; provenance and claimed scale remain unverified. |
| 7 October | German automotive client, document and financial data release asserted [S1] | No files or current access verified; compare a relevant alert against owned records before containment. |
| 1 November | German manufacturer appears in a forum data-leak listing [S1] | Sparse claim, without verified theft date or mechanism; do not count another intrusion from a forum repost. |
| 5 November | International travel/reservation dataset offered, with Germany among customer countries [S1] | The source organisation and a German organisation segment are unverified; no German provider intrusion inferred. |
| 8 November | Research/recruitment cloud-tenant administrator access advertised [S1] | Domestic ownership, customer geography and usable access unverified; review only a validated owned tenant. |
| 10 November | German forum/hosting-platform administrator access alleged [S1] | Customer and order categories are claimed; no active access or underlying platform intrusion is validated. |
| 17 November | Multi-country banking leads include a German segment [S1] | A lead list does not establish a bank intrusion; freshness and sale assertions remain unverified. |
These advertisements can concern different offers or overlapping older material. No shared operator, separate underlying intrusion for every row, or match to a public disclosure is established. The caller-ID and financial-contact rows support trusted verification routes, not claims about completed German fraud.
4.6 Validation questions for common claim types
| Claim type | What needs validation | What not to infer |
|---|---|---|
| Customer database | Origin, categories, age and ownership | Every claimed row represents a different person |
| Employee access | Account ownership, status and privilege | Domain association establishes employment |
| Remote access | A currently usable access path | An old screenshot establishes continuing access |
| Credential file | Freshness and previous remediation | Every line is valid or unique |
| Infrastructure inventory | Authenticity and current operational relevance | A document establishes control of the infrastructure |
| DDoS claim | Service state, duration and affected function | A momentary failed check establishes national disruption |
The organisation's incident record should preserve the proposition being tested, the source date, the owned asset or data category and the result. Keep alternatives visible. Assign the next action to an owner: identity for an active account, data protection for personal records, and the service operator for disruption.
This edition publishes no national asking-price median and estimates no completed sales. The quoted excerpts preserve the original language, with identifying material removed. Their translations convey the claim without adding technical or attribution certainty. [S1]
4.7 Publication platforms and languages
- Observed Germany-associated posts
Platform counts total 1,221,942; seven of the 1,221,949 observed posts remain unassigned in this breakdown. Source: Brandefense observed Germany-associated posts, 1 January-22 November 2026; cutoff day partial. Posts, retained claims and incidents are separate populations [S1].
| Platform label | Observed posts |
|---|---|
| Telegram | 1,218,904 |
| Forum | 2,539 |
| Discord | 197 |
| Paste services | 302 |
| Tor services | 0 |
The platform rows total 1,221,942 posts, seven fewer than the overall post total. The available aggregates do not establish the platform assignment of that difference. These rows are not a fully reconciled breakdown. [S1]
These are observed Germany-associated posts in the requested window. The platform labels describe publication venues, not verified attacks, successful transactions or domestic victim populations. They are not a complete catalogue of every underground venue. [S1]
- Observed Germany-associated posts
Language labels can overlap and do not establish nationality. Source: Brandefense observed Germany-associated posts, 1 January-22 November 2026; cutoff day partial. Posts, retained claims and incidents are separate populations [S1].
| Language association | Observed posts |
|---|---|
| German | 44,217 |
| English | 311,183 |
| Russian | 82,205 |
| Turkish | 788,012 |
| Ukrainian | 3,782 |
| French | 856 |
| Arabic | 7,994 |
| Chinese | 10,884 |
Language associations can overlap. They do not establish the nationality or residence of authors, customers or victims. Publication volume does not establish a national crime-family ranking or the frequency of successful intrusions. [S1]
4.8 Closing-period trade claims
The hosting offer disclaims live access. A 22 November offer advertises material associated with 16 websites and 10 small-business clients of a German hosting service. The seller expressly disclaims live SSH or root access. Possession, provenance, acquisition date and a completed sale remain unverified. Validate any relevant owned files and accounts before treating the offer as evidence of current access. [S1]
The credential posts repeat an advertising claim. Two 22 November posts advertise the same claimed 23,000-entry Germany credential list with matching cloud-access promotion. Their wording does not establish that the underlying records are identical, valid, current or independently stolen. Do not add the advertised counts or count the posts as separate breaches. [S1]
A 12 November offer advertises German investment and recovery leads. An 18 November offer labels German phone-contact material as 2023 data. Neither establishes the source breach, present accuracy or a new 2026 intrusion. Customer-protection teams should authenticate recovery contacts and payment requests through independent channels. [S1]
Key takeaways
- 01
Six illustrated claims show different validation tasks, not six confirmed intrusions.
- 02
Repeated advertising is not independent confirmation.
- 03
Route access, document and identity claims to the owner who can validate the specific proposition.
Chapter 5
Confirmed incidents and targeting by sector
This chapter answers
Which disclosures support a concrete defensive response, and which outcomes remain unresolved?
In short
Public disclosures describe different failures: stolen credentials, copied administrative data, supplier-held customer records and interrupted services. The cases below retain each source's limits. A shared sector or nearby date does not establish a common campaign. Start with the affected function and the confirmed access, then choose containment, customer support and continuity measures that fit that evidence.
5.1 How to read the cases
A confirmed security incident does not confirm every subsequent leak claim. We record the organisation's or authority's statement, identify the remaining uncertainty, and give a proportionate response. A case can include several phases of one incident, such as intrusion followed by patient extortion.
Named cases require official confirmation and wider reporting. Other cases use broad roles. Source titles and article addresses that would restore a withheld name are shortened under the publication policy. The private register retains the exact supporting passage.
This chapter is a set of documented cases, not an exhaustive national count. Discovery, disclosure, publication and restoration dates are kept distinct. Follow-up fraud and a new intrusion are not automatically counted as two separate compromises.
5.2 Public administration: data review after service containment
Case A
Berlin administration data loss
- Confirmed
- The Berlin data-protection authority identifies copying from administration networks between 7 and 12 August. It says employee information and possibly residents' information were published. The Senate's information page separately describes the continuing review. [S9][S10]
- Boundary
- An archive's public appearance does not establish that every file is authentic, that every person's data was published, or that access remains active. The reported incident must remain separate from any attacker's quantity or attribution.
- Interpretation
- Unstructured shared folders make scoping a distinct response task. Restoring a system does not identify every person whose correspondence may have been copied.
- Response
- Inventory the affected document stores, record which categories can be confirmed, and maintain a contact route for people awaiting an individual assessment.
The case supports a document-governance decision beyond network recovery. An organisation should know which shared stores contain identity documents, contractual records or sensitive operational plans. The data owner should be able to identify retention rules and who can approve access.
For communications, use a staged notice. Distinguish the known event from the still-developing assessment of individual impact. Give the next update time, the responsible contact route and advice matched to the confirmed categories. Do not promise that publication equals misuse, or that restoration eliminates misuse risk.
Case B
A state property agency's unresolved November incident
- Confirmed
- A state ministry reported an IT security incident on 19 November. Telephone availability was restricted and email was unavailable. The cause, extent and possible data outflow remained under investigation. [S22]
- Boundary
- The statement does not establish ransomware, copied data or an actor. A communications restriction is the confirmed operational consequence.
- Response
- Prepare alternate contact arrangements and preserve the investigation timeline. Update the public description when evidence establishes a more specific outcome.
This is a useful limit case. A public statement can responsibly confirm an incident before its mechanism is known. Defenders should not wait for an actor label to activate a fallback channel; analysts should not supply a label that the authority has not established.
5.3 Education: identity exposure and data access are different events
Case C
TUM campus credentials
- Confirmed
- TUM reported stolen central-login passwords, including an administrator's credential, and access against TUMonline. Its 9 November update restored the service with password changes and two-factor authentication required. [S11]
- Boundary
- The disclosure does not establish a verified theft total for the university's entire population. The wider reporting supports naming the organisation, not expanding the official impact statement. [S12]
- Interpretation
- An administrator credential can turn a portal event into a wider identity response. The response must address privileges and sessions alongside the original portal.
- Response
- Validate the affected accounts, revoke relevant sessions and review administrative changes before restoring normal access.
The recovery decision should be owned jointly by identity, application and service teams. Restoring the application's availability is only one milestone. Safe reopening also requires checking privileged grants, recovery methods and the audit coverage needed to investigate subsequent activity.
Use a synthetic recovery exercise to test the sequence. The exercise should demonstrate that a locked account can recover through a trusted route, that unauthorised sessions lose access, and that legitimate users can reach critical functions. Record exceptions rather than treating all users as identical.
Case D
Universität Tübingen employee records
- Confirmed
- On 10 November the university disclosed about 3,900 employee records, including identifying, contact and workplace information. It said students and passwords were not affected. It described a combination of recommended settings that exposed internal functions. [S13][S15]
- Boundary
- The disclosure does not identify a common operator with other university cases. Knowing a username was part of the described access condition; it does not make the username itself a secret.
- Interpretation
- Configuration combinations can create exposure even where each individual setting appears routine. Review the effective permissions of the whole service.
- Response
- Test authorisation at internal functions using an approved account, review supplier configuration guidance and warn affected employees about contextual phishing.
Case E
Universität Tübingen webmail notice
- Confirmed
- A separate 11 November notice described an overnight webmail attack and precautionary account restrictions. The university warned that passwords entered during the stated period might have been captured and initiated resets. [S14][S48]
- Boundary
- This report does not merge that notice with the employee-record event. A common institution and nearby date do not establish the same entry route, dataset or perpetrator.
- Response
- Match reset scope to the service and time interval described. Review linked services and sessions for affected accounts without asserting that every mailbox was read.
The two Tübingen cases illustrate why incident units matter. One notice concerns employee records and explicitly excludes passwords. The other concerns possible password capture in a particular mail-service interval. A combined headline would erase information needed to choose the right response.
Case F
LMU registration data
- Confirmed
- LMU's 19 September disclosure describes unauthorised access to registration records and says it must assume retrieval. The source lists identity, contact and some financial or administrative categories, depending on the individual record. [S49]
- Boundary
- The notice does not establish a public leak or a verified national student population. It distinguishes the affected registration material from examination and academic-performance information.
- Response
- Identify the affected data categories, support affected people and test authorisation on registration workflows before reopening them.
Across these education cases, the evidence supports several controls rather than one assumed initial-access pattern. Identity recovery, administrative privilege, application authorisation and personal-data scoping each need an owner. The report does not infer that all cases began with phishing or a shared software vulnerability.
5.4 Healthcare: protect patients during the secondary phase
Case G
Hospital intrusion followed by patient extortion
- Confirmed
- A hospital reported communications failure in February and a move to manual ward procedures. In March, the state data-protection authority and police warned that stolen patient information was being used in direct extortion messages. [S16]
- Boundary
- The later messages are a follow-on phase, not evidence of a second hospital intrusion. Neither a patient total nor a named operator is established here.
- Interpretation
- Sensitive information can reach the individual as an extortion demand after the institution begins recovery. The institution's response should include patient-facing support.
- Response
- Prepare scripts for recipients of extortion messages, preserve reports safely and explain that a payment does not establish deletion.
The lesson is organisational. A service desk cannot investigate a medical-data incident by asking patients to send sensitive records to an unverified address. It needs an approved route, clear triage questions and a way to connect urgent concerns to the incident team.
Clinical continuity and personal-data response should be exercised together. Manual procedures need staffing, reliable contact details and a safe way to reconcile records later. The organisation should specify who can resume the normal workflow and what evidence supports that decision.
Case H
Fresenius Medical Care's limited-system disclosure
- Confirmed
- The company reported unauthorised access to a limited number of internal systems on 22 September. It said medical devices, patient care, manufacturing and business continuity were unaffected. [S19][S20]
- Boundary
- This establishes the company's stated incident scope. It does not confirm an extortion group's claimed dataset, a German patient count or disruption of clinical devices.
- Response
- Separate internal-system investigation from clinical assurance. Validate service impact using the relevant operational owners rather than a criminal listing alone.
These hospital and medical-technology disclosures should not be collapsed into a general claim that healthcare systems were unavailable. One reports a manual operational fallback and later patient extortion. The other expressly states that medical care and production were unaffected.
Case I
Pharmaceutical cooling controls
- Confirmed
- LKA NRW's 18 November warning cites BSI reporting of attacks on a German pharmaceutical wholesaler's cooling systems. It separately discusses a French incident and states that direct effects on German blood banks had not been established. [S21]
- Boundary
- No common actor or common technical cause is established between the reported cases. The warning does not justify claiming German blood-product loss.
- Response
- Validate remote-maintenance access, independent temperature alarms and contingency storage. A backup of IT data does not maintain physical temperature conditions.
Chapter 6 develops the operational response. The immediate decision is to identify who receives an alarm, who can isolate remote access and who can protect stock while the system is investigated.
5.5 Supplier-held records and customer services
Case J
Photo-service supplier warning
- Confirmed
- A state data-protection authority warned in May about a supplier attack affecting photo studios. Children's photographs and personal information were potentially involved. [S17]
- Boundary
- Potential exposure is not a verified count of affected children. A common supplier does not establish a separate intrusion at every studio.
- Response
- Identify each controller's affected records and coordinate an accessible notice route. Confirm what deletion means across the supplier relationship before promising removal.
Case K
Hotel-software data and fraudulent messages
- Confirmed
- A July state-authority notice described another wave of hotel notifications after a software-supplier attack. Reservation and customer information were involved; the notice reported follow-up fraudulent messages. [S18]
- Boundary
- The statement did not establish bank or payment-data loss at that stage. A convincing payment demand can exploit reservation context without containing stolen card data.
- Response
- Brief reservations staff, publish the real payment process and provide an independent way for guests to verify a request.
The common response is to preserve useful context without disclosing additional information to the sender. A hotel should not authenticate a caller merely because the caller knows a booking date. A studio should not assume a record was erased everywhere because a user interface no longer displays it.
Contracts should identify how a provider will supply an affected-data list, preserve relevant logs and assist the controller's notification decision. The completion evidence is a rehearsed exchange of information, not a contractual promise whose practical timing has never been tested.
5.6 Transport and communications
Case L
Deutsche Bahn's February booking disruption
- Confirmed
- Deutsche Bahn's official reporting describes February overload of booking and information systems. Contemporary reporting describes attack waves on 17–18 February. [S7][S8]
- Boundary
- This is an availability case. It does not establish theft of passenger records, access to train-control systems or an official perpetrator attribution.
- Response
- Test customer information and ticket-access fallbacks alongside mitigation. Monitor the customer's completed task, rather than treating a responding homepage as full service restoration.
The incident illustrates a useful continuity measure: whether a passenger can retrieve a ticket or obtain reliable information during degradation. A network team's traffic chart is necessary evidence for mitigation, but it is not the same as a restored booking journey.
For communications services, distinguish an upstream dependency from an intrusion into every downstream organisation. Confirm which services are unavailable, provide a fallback contact route and record the dependency when assessing sector reporting duties.
5.7 Recent activity: 8–22 November
The source records 7,111 Germany-associated posts dated 8–22 November. That is observed publication volume, not a count of new incidents or verified German victims. [S1]
The closing fortnight includes new incidents, updates to earlier incidents and enforcement announcements. The table separates them so that publication dates do not create an artificial late-year surge.
| Date | Item | Evidence status | Immediate decision |
|---|---|---|---|
| 8–17 November | Reservation-provider access and rebuilt-service restart | One supplier incident, with fraudulent guest messages [S58] | Coordinate guest warnings with technical recovery |
| 8–20 November | Utility IT and customer-service disruption | Supply maintained; updates to one incident [S59] | Separate billing fallback from energy operation |
| 8 November | Banking-related CRM access offer | Unverified administrator-tenant advertisement [S1] | Validate account and tenant ownership |
| 9 November | Campus system recovery update | Update to TUM's 6 November disclosure [S11] | Validate reopening and account recovery |
| 9 November | Fake-shop charge announced | Alleged conduct from 2023–early 2025 [S42] | Treat it as enforcement context |
| 10 November | University employee records | Official data-access disclosure [S13] | Scope categories and warn staff |
| 11 November | University webmail | Separate password-risk notice [S14] | Match resets to stated exposure |
| 12 November | Malware disruption announced | Police action on 9–11 November [S23] | Act on relevant infection notifications |
| 12 November | Social-service communications restrictions | Municipal confirmation; in-person work continues [S55] | Publish working contact routes |
| 12 November | Defence supplier statement | Incident acknowledged; further details unresolved [S50][S86] | Verify supplier updates through a trusted route |
| 13 November | Political-party personal-data notice | September–October activity; recent notification [S56] | Scope records without assigning political motive |
| 13–20 November | Hosting-provider access and scope updates | Later higher-risk instance notice [S54] | Make a tenant-specific determination |
| 15–18 November | County hospital incident and reopening | One incident, several recovery milestones [S53] | Coordinate clinical diversion and reopening |
| 16 November | German-organisation file offer | Unverified document advertisement [S1] | Validate origin and age through owned records |
| 17 November | Bank response to document reporting | Statement disputes a current intrusion [S39][S40] | Preserve the distinction from historical disclosure |
| 17 November | Municipal administration | Police confirm unauthorised access; theft still checked [S51] | Preserve logs and provide alternate contacts |
| 17–20 November | Local communications attack and recovery | Provider confirmation and service update [S52] | Test customer support and dependent services |
| 18 November | Industrial cooling warning | Official operational-security warning [S21] | Test independent monitoring and maintenance access |
| 19 November | State property agency | Security incident confirmed, possible outflow unresolved [S22] | Activate communications fallback |
| 19–20 November | BSI annual assessment | New publication covering an earlier window [S3] | Update planning without re-dating the events |
| 22 November | Hosting data-lot offer | Unverified data offer; seller disclaims live access [S1] | Validate owned files and accounts before containment |
| 22 November | Matching credential-list advertisements | Two posts repeat an unverified advertised claim [S1] | Check account ownership, age and prior remediation |
The table is not a complete list of fortnight incidents. Some late discoveries remain unresolved in the coverage register. This report makes no claim that missing recent posts or notices imply a reduction in threat activity.
5.8 Other confirmed closing-period cases
Case M
Municipal access under investigation
- Confirmed
- A 17 November police notice described unauthorised access to part of a municipal administration's IT, disruption and disconnection of systems. Investigators were checking possible data theft. [S51]
- Boundary
- A message requesting contact does not establish encryption, a specific ransom demand or copied data.
- Response
- Preserve logs and the message, arrange alternative citizen contacts and update the scope as evidence develops.
Case N
A local communications provider's outage
- Confirmed
- A provider attributed telephone and internet disruption overnight on 17–18 November to a cyberattack. Its 20 November update reported stable services while customer-service telephone availability was being restored. [S52]
- Boundary
- The notice did not establish data theft or a named actor. Restoration of network service and restoration of support channels are distinct milestones.
- Response
- Test alternate customer contact routes and track restoration by function, including dependent organisations.
Case O
A county hospital's emergency-service interruption
- Confirmed
- A county reported an attack detected on 15 November and temporary withdrawal of emergency admission. Later notices reported emergency admission restored on 17 November and functioning operating theatres on 18 November. [S53]
- Boundary
- These are phases of one incident. The notices do not establish a named actor or data theft.
- Response
- Rehearse diversion, manual operation and safe reopening with clinical leaders. Track postponed work after the technical recovery milestone.
Case P
Hosting-provider infrastructure access
- Confirmed
- A hosting provider described unauthorised infrastructure access on 13–14 November, revoked access and rotated credentials. Its 20 November update identified a small higher-risk instance population for direct contact. [S54]
- Boundary
- Earlier statements of no indication of personal-data leakage do not establish no customer impact. The later update does not establish universal customer access either.
- Response
- Follow trusted provider guidance, review your own access evidence and record a tenant-specific determination.
Case Q
A social-service provider's restricted communications
- Confirmed
- A municipal notice on 12 November described an attack affecting part of a social-service provider's IT and restricted email. In-person services continued, with telephone or on-site contact available. [S55]
- Boundary
- No data outflow was known at that point. That is a time-bounded assessment, not proof of absence.
- Response
- Publish the working contact routes and preserve continuity for people who depend on the service.
Case R
A political party's historical incident notice
- Confirmed
- AfD's 13 November notification described unauthorised access to mailboxes in September and further systems in October. It reported copied personal data and continuing investigation. Earlier press reporting also covered the incident. [S56][S57]
- Boundary
- The notice date does not make the underlying activity a new November intrusion. No verified affected-person total or actor attribution is established here.
- Response
- Separate externally hosted mail from party administration systems, scope the affected records and preserve notification decisions. Apply the same privacy and response standards regardless of affiliation.
No judgement about the party's policies, government decisions or election outcomes follows from this incident. The case is included for its documented security and personal-data consequences.
Case Y
Reservation-provider access followed by fraudulent messages
- Confirmed
- A reservation software provider says an attacker entered through a customer account on 8 November and retrieved guest data on 8–9 November. It reports subsequent fraudulent emails seeking payments or card information. It took the application offline on 13 November because deeper access could not be excluded and restarted a rebuilt service on 17 November. [S58]
- Boundary
- The exact data volume remained under investigation. These are phases of one supplier incident; downstream restaurant warnings are not separate confirmed restaurant intrusions. No common actor with earlier hotel-software incidents is established.
- Response
- Obtain a scoped provider assessment, warn guests through a trusted channel and authenticate payment requests independently. Service restoration does not close the customer-fraud response.
Case Z
Utility IT disruption with energy supply maintained
- Confirmed
- A municipal utility says it detected an attack on 8 November and isolated affected systems. Customer portals, email, meter appointments and several payment or collection processes were affected. Updates through 20 November said electricity, gas and district-heating supply remained uninterrupted. [S59]
- Boundary
- The provider had no indication of personal-data outflow at the stated time. That is a qualified assessment. Neither an energy-supply outage, a specific actor nor an entry route is established.
- Response
- Test customer-service and billing fallback separately from energy operation. Record delayed payments and appointments so that restoration includes the accumulated customer work.
5.9 Earlier incidents and important exclusions
The record extends beyond the closing fortnight. The following notices add confirmed events and explicit exclusions. They preserve the distinction between an event, an update and a precautionary shutdown.
| Period | Role and supported finding | Limit and response |
|---|---|---|
| 21 January; 19 February notice | Police and prosecutors confirm an attack on a state cultural collection's IT [S60] | Route, actor and data quantity were not disclosed; preserve the investigation boundary |
| January; restored 23 January | A professional chamber removed website malware and warned that dashboard passwords might have been intercepted [S61] | Possible interception is not confirmed for every user; no pharmacy-care system impact is established |
| 16–24 April | A municipality confirmed ransomware and system isolation, followed by restored registry, civil-status and payment functions [S62] | The updates concern one incident; personal-data theft remained unresolved |
| April event; 22 May notice | A university hospital confirms that patient information was stolen at its external billing provider, while its own systems and patient care were unaffected [S75] | Initial disclosure only; later scope remains unresolved here, so no final patient total is published |
| Learned 10 June; 25 June notice | A research institute confirms access to file servers and stored files, followed by disconnection and resets [S65] | Which files were actually taken remained unresolved; no actor is established from its research subject |
| Detected 19 June; updated 31 July | A state museum confirms unlawful copying involving staff, partner and visitor information [S71] | No final affected-person count or actor; scope data categories and affected-person support |
| 28 June–16 July | A municipal utility confirms disruption to portals, communications and payments, while maintaining supply [S63] | Its later assessment excluded loss of current customer data but left older customer-related data unresolved; one incident across recovery updates |
| 16 July; 21 July statement | An official-origin municipal statement reports business-contact and technical-directory data escaping after an attack on an IT provider [S64] | It says citizen, private staff-contact and specialist-application data were unaffected; the republished statement is not independent editorial confirmation |
| July; August updates | A research institute reports an attack on a network component and precautionary communications disruption [S66] | Its investigation said internal-system access attempts failed; do not convert an outage into confirmed internal access |
| August and October; 8 October notice | A memorial foundation explicitly distinguishes a renewed website attack from its earlier IT attack [S67] | Two events, no established common actor or political motive; the October notice supplies the distinction |
| 27 September; 29 September notice | A county describes preventive shared-system shutdown and patching [S73] | It expressly reports no successful attack or detected outflow; excluded from confirmed intrusions |
| 15 October notice | A municipality confirms copied data and manual recovery processes, with data scope still under investigation [S68] | Its utility was expressly excluded; do not extend the incident to every affiliated entity |
The January chamber and November university cases both raise password questions, but there is no evidence of a shared campaign. The two utility cases support separating supply continuity from customer operations, not a claim of one utility-sector actor. A research institution's subject area and a memorial's public role do not establish motive or state attribution.
A separate university-hospital notice on 22 May also confirms patient data lost at an external billing provider while the hospital's own systems and care remained unaffected. Its affected-client population is not added to another hospital's figures. The two client disclosures alone do not establish an additional provider incident or a combined patient total. [S83]
5.10 Further priority-organisation findings
Development supply chain. A vendor warned that four malicious open-source package versions were distributed through a public package registry on 29 April. It warns of credential extraction and attempted propagation after installation. This establishes a software-supply-chain incident, without a verified German downstream victim count. Development teams should identify affected builds, follow the vendor's mitigation and assess credentials reachable from those environments. [S76]
University maintenance and uncertainty. A university's 25 February notice describes emergency maintenance on 23 February after an external vulnerability report. It says student and former-student addresses might have been accessible, but exploitation could not be determined conclusively. The patch and interruption are confirmed; a successful intrusion is not. This belongs in the security-event record with that qualification. [S77]
Vehicle-retail supplier. A retailer's October notice describes unauthorised access to a shared provider during 8–30 September and possible customer, vehicle and purchase information exposure. It distinguishes confirmed client-record access from records whose access could not be excluded. Its statement that bank data were unaffected applies to its notice; it is not a blanket exclusion for every provider customer. Protect invoice verification and obtain a client-specific scope. [S80]
Certification services. A certification organisation acknowledges a September IT security incident in an October publication. The supporting passage does not establish a data category, actor or affected-client count. It is retained as a bounded service-provider notice, without assuming that certification customers' systems were accessed. [S84]
The Berlin case also has a recent federal-scope update. The 16 November parliamentary summary warns that names or abbreviations in file names can misidentify an affected authority, while other affected material may be missed by file names alone. Content review is required before assigning federal impact. This is a scope update to Case A, not another attack. [S82]
5.11 What the cases support
| Evidence pattern | Supported interpretation | First response |
|---|---|---|
| Administrative records copied | Scoping can continue after containment | Assign data owners and affected-person support |
| Credentials used across services | Identity can extend the incident's reach | Review privileges, sessions and recovery methods |
| Supplier-held records exposed | Customer harm can occur outside the controller's own systems | Obtain a scoped record list and coordinate notices |
| Availability disrupted | Business tasks can fail without proven theft | Test the minimum service and alternate channel |
| Industrial control targeted | Operational safety requires physical verification | Validate independent alarms and safe fallback |
The cases do not establish a national sector ranking. Public disclosure is uneven, and some incidents remain private. Their value lies in showing the response questions that similar organisations should be able to answer.
Key takeaways
- 01
Preserve the distinction between credential misuse, copied data and unavailable services.
- 02
The two Tübingen notices describe separate affected populations and responses.
- 03
Recent publications include historical enforcement and recovery updates as well as new incidents.
Chapter 6
Hacktivism, DDoS and operational technology
This chapter answers
What distinguishes a public availability claim from confirmed operational impact?
In short
An availability claim can circulate before an operator understands the outage. A confirmed outage still does not establish data theft, an actor or physical-process access. Germany's railway case and November cooling warning illustrate different response needs. Prepare mitigation and public status messages, but verify operational consequences with the people responsible for the service and equipment.
6.1 Evidence timeline
- 6 Feb 2026BSI and BfV
Messenger-phishing warning; identity campaign, not a DDoS event [S5]
- 17 Feb 2026
- 20 Mar 2026BKA and ZAC NRW
Announce action against infrastructure associated with Aisuru and Kimwolf [S24]
- 31 Aug 2026State criminal police
Announce a DDoS and attempted-extortion investigation; allegations remain under investigation [S25]
- 12 Nov 2026BKA and ZIT
Announce malware-infrastructure disruption during 9–11 November [S23]
- 18 Nov 2026LKA NRW
Warn about attacked cooling controls and operational contingency planning [S21]
Every row identifies its source type. The identity warning and malware enforcement rows are contextual events, not additional DDoS incidents. Their inclusion does not establish a shared campaign or operator.
6.2 Availability is a customer outcome
The February railway disclosure establishes service disruption. It does not attribute the event to a named claimant. The report therefore does not convert unrelated actor rhetoric, a foreign-policy theme or an availability-check result into official attribution. [S7][S8]
For a customer-facing service, test the complete task. A homepage may respond while authentication, ticket retrieval or payment fails. Record which functions work, from which user groups, and which fallback route remains usable. Keep the measurement date beside the public status.
The response should distinguish traffic mitigation from business restoration. Infrastructure teams can confirm that traffic is being absorbed. The service owner confirms that the essential user journey works. Communications should state the narrower fact until both are established.
| Question | A claimant may assert | Established reporting | Defender's test |
|---|---|---|---|
| Is the site down? | A momentary failed check | Railway booking disruption was reported [S7][S8] | Test the essential customer task |
| Was data taken? | An outage presented as access | The cited availability case does not establish data theft | Review authenticated access and data-access evidence |
| Who caused it? | A self-identified actor takes credit | No official actor attribution established in this case | Preserve evidence and state attribution separately |
| Was an industrial process altered? | A claimed control-panel screenshot | The cooling warning identifies a distinct operational risk [S21] | Verify process readings with the operator |
The claimant column is a general interpretation aid, not a quotation from a German post. It must not be read as evidence that a particular actor made each assertion during this window.
6.3 Criminal disruption is not always political
The August state-police notice describes a DDoS and attempted-extortion investigation. The March BKA/ZAC notice concerns botnet infrastructure. These official accounts support attention to criminal infrastructure and incentives; they do not prove that a specific German outage used those botnets. [S24][S25]
The BKA's retrospective discussion of NoName057(16) concerns 2025. Its assessment supplies historical context for public-service readiness. It cannot establish the group's responsibility for the February 2026 railway incident. [S4]
Treat an announced political cause as the speaker's framing. Response teams should assess traffic, impact and access evidence on their merits. The same protection and notification standards should apply to political organisations, government bodies, businesses and civil society.
ENISA's 4 November joint statement describes refrigeration incidents in several EU member states, including a further wave on 1–2 November. It gives no German victim count. Its regional scope supports reviewing exposed control systems, but it does not add German blood-bank incidents to the national record. [S70]
6.4 Industrial cooling changes the continuity question
The November warning adds a physical-process dimension. Its German and French examples remain separate. The source does not establish direct impact on German blood banks or identify an operator common to the reported events. [S21]
The practical question is whether staff can detect an unsafe condition independently of the controller under investigation. A screen showing a normal temperature is insufficient if the same system supplies the reading and the alarm. Validate an independent monitoring route with the equipment owner.
The response plan should specify who can isolate remote maintenance without creating a safety problem. It should identify alternative storage, escalation contacts and the maximum acceptable interruption for the relevant stock. Those decisions require facilities, operations and safety staff alongside information security.
Do not apply a generic shutdown instruction to every industrial system. Unplanned isolation can interrupt monitoring or a necessary safety function. Use the approved operational procedure and retain an authorised record of changes and physical observations.
An August state-security warning about programmable controllers primarily concerns attacks reported against US facilities. It recommends hardening, separation and monitoring, but provides no German victim count. The warning supports reviewing comparable equipment; it does not establish an intrusion into a German manufacturer or identify an actor in this report. [S79]
6.5 A claim-to-response decision sequence
- Establish the service state. Confirm the affected function, duration and available fallback. Record customer or operational impact separately from a network symptom.
- Check access evidence. Review whether there is evidence of unauthorised identity use or data access. Do not infer either from an outage alone.
- Verify the physical process. Where industrial equipment is implicated, use independent readings and the operational owner. Preserve the distinction between a screenshot and confirmed manipulation.
- Publish a bounded statement. Identify the known disruption, the action underway and the next update time. State what remains under investigation without promising an absence of impact.
- Retain the evidence for reporting. Preserve the chronology and contact the competent route under the organisation's obligations. Record decisions even when no formal notification is required.
6.6 Completion evidence
| Owner | Test | Completion evidence |
|---|---|---|
| Infrastructure | Activate contracted DDoS mitigation | Exercise log and escalation contact |
| Service owner | Complete the minimum user journey during degradation | Test record and fallback instructions |
| Communications | Issue a claim-aware service statement | Approved template and update schedule |
| OT and facilities | Detect an abnormal condition independently | Witnessed alarm and response test |
| Supplier management | Restrict maintenance access under the approved procedure | Named access owner and tested revocation |
Key takeaways
- 01
A failed availability check does not establish theft, attribution or sustained outage.
- 02
Political framing and criminal infrastructure are separate propositions.
- 03
Industrial incidents require physical verification and an operational fallback alongside IT recovery.
Chapter 7
Threat actors and evidence relationships
This chapter answers
Which actor-related propositions are supported, and which relationships remain unestablished?
In short
Extortion listings provide the strongest comparable actor population in this edition. Qilin, SafePay and The Gentlemen lead, but most German listings carry other labels. Official campaign warnings add identity and espionage context without establishing a German victim in every case. Keep listing counts, state attribution and shared-operator hypotheses separate. Build response plans around the affected access and service rather than an actor's reputation.
7.1 Relationship map
- Own post
- Relay
- Official statement
- Public reporting
- Our observation
No evidenced link to other actors
Each edge identifies a specific relationship: a listing label in the German population, a campaign warning or an official enforcement statement. An edge does not establish a common operator. Unconnected actors remain unconnected; a shared country or technique is not enough to join them.
The diagram is not a map of the entire German threat environment. Its purpose is to make the report's evidence relationships legible. A new claim must establish its own connection before it is added to an existing campaign.
7.2 Ransomware and extortion
- 1 January-22 November 2026
- Full year 2025
Unmatched periods and different country populations; not a national risk ranking. Source: Brandefense Ransompedia Germany-labelled listings, 1 January-22 November 2026; November partial. Listings record criminal claims, not confirmed intrusions [S2].
Comparison countries were selected as examples to provide context only; they do not imply a ranking or political assessment.
The five country counts are shown together for context. They are not adjusted for economic size, organisation population, disclosure behaviour or source coverage. They cannot establish which country's organisations are safer.
| Country | Requested 2026 window | Full-year 2025 | Largest current label |
|---|---|---|---|
| Germany | 457 | 311 | Qilin, 62 |
| France | 295 | 182 | Qilin, 48 |
| Netherlands | 77 | 44 | LockBit5, 15 |
| Poland | 56 | 32 | The Gentlemen, 19 |
| Austria | 61 | 38 | The Gentlemen, 11 |
Source: Ransompedia [S2]. Every current-window figure is compared with a full prior year, not the same elapsed period. No growth-rate claim is made.
Qilin
In Germany, 2026
- 2026
62 German listings in the requested window
Claimed targets
Sector mix in the listing population; no named victims here
Seen in public reporting
- [S2] Ransompedia records 62 German listings under this label.
Our assessment
Listing presence supports recovery and disclosure readiness. It does not establish encryption, theft, a shared operator or a completed payment.
ConfidenceHigh for retained listing count; underlying assertions remain claims
SafePay
In Germany, 2026
- 2026
55 German listings in the requested window
Claimed targets
Sector mix in the listing population; no named victims here
Seen in public reporting
- [S2] Ransompedia records 55 German listings under this label.
Our assessment
Listing presence supports recovery and disclosure readiness. It does not establish encryption, theft, a shared operator or a completed payment.
ConfidenceHigh for retained listing count; underlying assertions remain claims
The Gentlemen
In Germany, 2026
- 2026
44 German listings in the requested window
Claimed targets
Sector mix in the listing population; no named victims here
Seen in public reporting
- [S2] Ransompedia records 44 German listings under this label.
Our assessment
Listing presence supports recovery and disclosure readiness. It does not establish encryption, theft, a shared operator or a completed payment.
ConfidenceHigh for retained listing count; underlying assertions remain claims
The leading German labels account for 62, 55 and 44 listings. Their combined 161 represents 35.2% of 457. The remaining 67 labels account for 296. A preparedness plan limited to the leading three would therefore omit most of the retained listing population. [S2]
This distribution supports capability-oriented planning. Test how identity access can reach critical data, how backups are protected from the production identity system, and how a data-loss notice is prepared. Those decisions remain useful when a claimant changes its name.
7.3 Monthly and sector composition
- Germany-labelled extortion listings
Source: Brandefense Ransompedia Germany-labelled listings, 1 January-22 November 2026; November partial. Listings record criminal claims, not confirmed intrusions [S2].
| Month | German listings | Month | German listings |
|---|---|---|---|
| January | 23 | July | 58 |
| February | 30 | August | 64 |
| March | 32 | September | 35 |
| April | 44 | October | 65 |
| May | 32 | November to cut-off | 33 |
| June | 41 | Total | 457 |
Source: Ransompedia [S2]. The counts describe publication timing. October's 65 listings do not establish 65 intrusions beginning that month. November is incomplete and is not directly comparable with a full month.
The largest retained industry label is manufacturing, with 77 listings. Consulting follows with 55. These labels describe the listing records; they are not a denominator for the number of organisations operating in each industry. [S2]
Manufacturing leaders should read this as a continuity-planning reason to test recovery and supplier access. Consulting and service providers should test whether access or data held for customers increases the consequences of an incident. The listing figures alone do not establish either entry route.
7.4 Messenger phishing and state-attributed campaigns
Unattributed messenger-phishing campaign
In Germany, 2026
- 2026
Official reporting context; dates in Chapter 7
Claimed targets
Only as described by the cited source
Seen in public reporting
Our assessment
Legitimate messenger features and social engineering are described. No software-encryption failure is established. Review linked devices and trusted recovery.
ConfidenceAs bounded in the cited official reporting
The messenger case is an official warning about social engineering of legitimate features. It is not evidence that the encryption protocol or application software was broken. Attribution remains as described by BSI, BfV and the authorities named in their reporting. [S5][S6]
For leadership teams, the defensive unit is the conversation and its participants. Review linked devices, recovery arrangements and trusted contact routes. Train staff to verify a security request through a known channel before approving an account action.
APT28
In Germany, 2026
- 2026
Official reporting context; dates in Chapter 7
Claimed targets
Only as described by the cited source
Seen in public reporting
- [S36] Official supporting passage read
Our assessment
The source warns about European policy-research targets. It does not establish a named German victim or connect unrelated German incidents.
ConfidenceAs bounded in the cited official reporting
A January state-security warning relays credential-phishing activity against European think tanks and associates it with APT28. It supports planning for policy and research organisations. It does not establish a named German victim or justify assigning an unrelated German incident to that group. [S36]
The maritime warning is similarly bounded. The authority discusses espionage in seven unspecified EU states and states its own attribution. Germany cannot be inserted into an unnamed country list merely because the warning is published by a German authority. [S38]
A 30 April state-security warning describes fake conferencing messages that lead users to execute a command and install macOS malware. The authority attributes that campaign to Lazarus. This is a campaign warning, not a confirmed German victim count. For business meetings, provide a known support route and train staff to stop when an invitation asks them to run an unexpected command. [S72]
7.5 Hiring and supplier identity
A state-security warning describes remote IT applicants using false identities and gives the authority's North Korea linkage. The defensible response is identity, qualification and access verification. It is not screening based on appearance, language or an assumed nationality. [S37]
Apply the same process to every supplier worker. Verify the contracting entity, employment relationship, access purpose and device arrangements. Review who can approve remote access and how it is revoked when the assignment ends.
These checks also help distinguish an authorised contractor from an unknown session. An organisation should be able to connect each privileged identity to a current accountable person and engagement without relying on a supplier's email introduction alone.
7.6 Enforcement and the limits of disruption
Malware named in Operation Endgame
In Germany, 2026
- 2026
Official reporting context; dates in Chapter 7
Claimed targets
Only as described by the cited source
Seen in public reporting
- [S23] Official supporting passage read
Our assessment
BKA and ZIT name Gootloader, CastleRAT and StealC in the November action. Global quantities are not German victim totals. Local infection validation remains necessary.
ConfidenceAs bounded in the cited official reporting
The November BKA/ZIT statement identifies Gootloader, CastleRAT and StealC in an operation conducted on 9–11 November. It reports global infrastructure and credential findings, including infrastructure in Germany. Those quantities are not German victim totals. [S23]
An infrastructure action does not demonstrate that every previously infected device has been cleaned. Organisations receiving a relevant notification should validate the host, credentials and active sessions through their own incident process. Preserve notification dates and the resulting disposition.
The March botnet and August DDoS investigations belong to different official case populations. This report does not connect them to Endgame, the railway disruption or the cooling incidents without a concrete shared element. Suspicions and enforcement announcements remain distinct from convictions. [S24][S25]
7.7 Vulnerability handling as a cross-sector control
The BSI's retrospective account of mobile-device-management exploitation describes vulnerable systems in Germany and distinguishes confirmed data outflow from cases where it could not be excluded. No German victim total is established in that account. [S45]
Separately, the following verified vulnerability is a control example, not evidence of exploitation at a German organisation in this report.
| Reference | Affected products | Verified remediation | Limit |
|---|---|---|---|
| CVE-2026-1731 | BeyondTrust Remote Support 25.3.1 and earlier; Privileged Remote Access 24.3.4 and earlier | Vendor patches BT26-02-RS / BT26-02-PRA; Remote Support 25.3.2 or later; use Privileged Remote Access 25.1.1 or later as the vendor's explicit upgrade guidance | Confirm supported upgrade path and patch eligibility with the vendor |
Sources: NVD record and vendor advisory [S33][S34]. Older unsupported releases need a supported upgrade path. Applying an update does not by itself establish whether an earlier intrusion occurred.
For an exposed remote-support service, the completion evidence should include the installed version, patch confirmation, access restrictions and an investigation decision. Do not claim a German exploitation campaign from a global vulnerability record.
Key takeaways
- 01
Most German listings sit outside the three leading claimant labels.
- 02
State attribution belongs to its named source and does not transfer to unrelated incidents.
- 03
Enforcement and patching require local validation of hosts, identities and continuing access.
Chapter 8
Fraud, identity and secondary harm
This chapter answers
How can information loss produce further harm after an incident is contained?
In short
Official warnings show stolen information being reused against patients and hotel guests. These cases support customer protection alongside technical recovery. Recent enforcement also concerns fake shops and investment-fraud contact infrastructure, but those quantities measure different things. Prepare a trusted verification route, brief customer-facing teams and record reports of follow-up contact without promising that the institution can prevent every attempted fraud.
8.1 A supported chain, with explicit limits
The patient-extortion and hotel-software notices establish that copied information was used in further contact with affected people. They do not establish the entire criminal supply chain, a completed sale or a common operator between the cases. [S16][S18]
The following is an analytical model for defensive planning. It connects data exposure, plausible impersonation, an induced action and monetisation. The model is not a finding that the same seller, caller and recipient account operated together in the German cases.
| Stage | Possible abuse | Defender's interruption |
|---|---|---|
| Information exposed | Real identity or reservation details supply credibility | Identify the affected categories and warn users proportionately |
| Contact initiated | A message imitates a familiar institution | Offer a separately verified contact route |
| Action requested | Recipient is asked to pay or disclose a code | Publish the legitimate process and brief staff |
| Value moved | Payment reaches a fraudulent destination | Escalate promptly to the payment provider and police |
| Repeat contact | Another message uses the same information | Retain incident context and update support guidance |
The business implication is that confidential information can remain useful after passwords have changed. A person cannot easily replace a medical history or a past reservation. Support should therefore address the information's sensitivity and context, not only authentication.
8.2 What customer operations should be able to answer
A recipient needs a simple way to determine whether a message is authentic. The organisation should identify its genuine payment and support processes, a route for independent verification and what information staff will need to assess a report.
The first conversation should avoid adding sensitive details to an untrusted exchange. Staff should not authenticate a caller solely because the caller quotes a booking, an address or a treatment. Those details may be the very information involved in the incident.
For affected people, the institution should distinguish recommended action from a guarantee. A warning can explain how to verify requests and preserve suspicious contact. It cannot establish that a criminal no longer holds the data or that paying an extortion demand causes deletion.
8.3 Police and regulator reporting on fraud
| Published | Official action | Unit and limitation |
|---|---|---|
| September | Authorities reported disabling 9,304 German telephone numbers associated with suspected investment-fraud activity over the preceding period [S43] | Telephone numbers; not victims or separate criminal groups |
| October | Investigators described misuse of a local party email account to send almost 5,000 messages [S41] | Messages; no political motive established at that stage |
| November | Prosecutors announced a fake-shop charge concerning alleged earlier conduct [S42] | Enforcement announcement; charge is not conviction |
The investment-fraud action illustrates an interruption point outside the victim organisation's perimeter. The account-misuse case illustrates that a familiar sender can become a fraud channel. Neither establishes a new national loss estimate.
The recent fake-shop announcement concerns alleged conduct from 2023 to early 2025. It belongs in the recent enforcement section because of the announcement date. It must not become a count of fake shops newly created in November 2026.
A separate May BKA/ZIT notice describes disabling a rebuilt illegal marketplace and arresting its alleged operator. It is an enforcement action, not a count of German victim organisations or evidence that every advertised product was delivered. [S47]
The Hessian data-protection authority also warns that authentic booking details can make fraudulent guest messages convincing. Its broader accommodation warning does not establish that payment data were taken in the specific November reservation-provider case. [S69]
8.4 Public authorities as a trusted verification route
A public body should explain its real process for payments, document requests and account recovery. The explanation must be findable without following a link from the suspicious message. Customer-facing staff should know how to redirect a concerned user to that process.
For an email-account incident, verify mailbox rules, delegated access and authorised applications alongside the password. Preserve the messages needed for investigation. Coordinate a recipient warning when there is evidence that an account was used to contact others.
For a bank or merchant, the relevant controls include new payment destinations, unusual account activity and customer reports of pressure to act. A leaked contact list alone does not establish that a transfer is fraudulent. The fraud process must validate the transaction and the customer's circumstances.
8.5 Card material in posts: a narrow view
Scope of the card figures
The figures describe payment-card records in the retained post corpus. They do not measure national card exposure, card theft or fraud loss.
The retained post material contains 15 distinct German-associated payment-card records across 16 posts. This is a narrow observation of published card material, not a German issuer census. Country association does not independently establish the cardholder's residence or current card usability. No issuer ranking, completed fraudulent transaction or loss total is inferred. [S1]
An issuer should validate an exposure through its established compromise-handling process. A post's assertion that a card is usable does not establish an authorised transaction, a completed fraudulent payment or a loss. Do not use the narrow post population to rank national payment security.
8.6 Prices and market size
This edition publishes no German national underground price index. Asking prices, where observed in a specific advertisement, are claims about an offer. They do not establish a completed transaction, the seller's possession of the material or the value of the underlying crime.
An estimate of national fraud proceeds would require evidence that this report does not hold. Repeated advertisements can reflect the same offer, and a message can name several countries. The appropriate action is validation of a relevant claim, not multiplication of posts by an asking price.
8.7 Completion evidence for customer protection
| Owner | Required capability | Evidence of completion |
|---|---|---|
| Customer operations | Recognise incident-related contact | Approved scripts and a staff exercise |
| Fraud | Assess suspicious payment requests | Documented escalation and transaction review |
| Data protection | Determine affected categories and people | Scoped assessment with remaining uncertainty |
| Communications | Provide independent verification | Published trusted contact route |
| Security | Link follow-up reports to the incident | Incident record with a dated disposition |
Key takeaways
- 01
Copied information can support later extortion or impersonation after systems recover.
- 02
Enforcement counts of numbers, messages and historical charges are not national fraud totals.
- 03
Give affected people a trusted verification route and trained support staff.
Chapter 9
Banking and financial services
This chapter answers
Which financial-domain observations require account and fraud validation?
In short
The selected financial portfolio has 25 observed domains, with current credential offers for 13. It records at least 80,163 device associations and 669,977 credential-record associations. These observations do not establish bank intrusions or unique customers. Validate current accounts, distinguish staff from other users and keep customer-fraud evidence separate from institution-access claims.
9.1 Financial-domain exposure observations
| Measurement | Result | Interpretation |
|---|---|---|
| Financial domains with observations | 25 | Selected services; no census of independent banks |
| Domains with current credential offers | 13 | Account-validation leads, not confirmed intrusions |
| Domains with device associations | 21 | Association does not establish device ownership |
| Device associations, at least | 80,163 | Overlapping cumulative domain associations |
| Employee-device associations | 698 | Validate employment and account context |
| Other device associations | 79,465 | Not verified customer devices |
| Credential-record associations | 669,977 | Not unique people or active credentials |
Exposure source: dated observations [S87]. The separate payment-card observations in §8.5 do not establish the source of card exposure. [S1]
Scope of the card figures
These figures describe selected financial-domain exposure observations. They do not establish independent banks, unique customers or confirmed intrusions.
No bank or sector rating is published. The measured counts above retain their separate denominators.
9.2 A recent bank statement is not a new-intrusion count
On 17 November, Deutsche Bank said the documents described in recent media reporting were historical. It stated that its checks had found no indication of current IT access or access to account information, logins or payment data. Its analysis continued. [S39]
Russian-language reporting covered the disclosure and regulatory attention. That supplies reporting context, not independent confirmation that the bank suffered a new intrusion during this report's window. This report takes no position on the unrelated political or client matters discussed in that reporting. [S40]
| Proposition | Media disclosure | Bank's statement | Report treatment |
|---|---|---|---|
| Documents exist | Reporting describes internal material [S40] | Bank describes it as historical [S39] | Document disclosure, with age unresolved at individual-file level |
| Current IT access | Not independently established here | Bank says no indication from its checks [S39] | Not counted as a confirmed new intrusion |
| Login or payment access | Not established by a document-volume headline | Bank says no indication [S39] | No exposure total inferred |
| Public interest reporting | Reporting examines the material | Bank states its position | No inference about a cyber operator or motive |
The correct defensive question is whether your own evidence shows a currently usable identity, session or access path. A document's publication date does not answer that question. Equally, an institution's statement of no indication remains its assessment at that time.
A 14 August BKA/ZIT announcement concerns an investigation into unauthorised online-banking debits allegedly made in November 2023 after a faulty provider update. It reports international enforcement activity in August 2026. The authority does not name the affected bank. This is historical alleged conduct with a recent enforcement phase, not a new German bank intrusion or a 2026 loss total. [S81]
9.3 Customer-side and institution-side evidence
If an alert names a bank login, determine whether the account belongs to a customer, employee, contractor or unknown person. A service-domain association alone does not establish a device owner or an intrusion into bank infrastructure.
For a validated employee identity, examine privilege, sessions and recent access. For a customer, combine the alert with account-security and fraud evidence. Apply the bank's established authentication and support processes before requesting a disruptive action.
The different owners should share a common disposition vocabulary: confirmed and contained, already remediated, unrelated, unresolved or escalated. The evidence should show why the case was closed. A dashboard count without dispositions does not demonstrate reduced risk.
9.4 Supplier and payment-process dependencies
The hotel and healthcare warnings show how accurate contextual information can make a fraudulent request plausible. Banks should prepare customer-support routes for requests that appear to quote real institutional details. This is a defensive inference from those cases, not a measurement of banking fraud prevalence.
For the institution itself, map payment operations to identity, communications and third-party services. Include a fallback channel for authenticating urgent instructions when normal email or telephony is unavailable. Test who can approve the exception and how the decision is recorded.
For a changed payee or supplier instruction, use a known independent contact route. Do not accept a new number supplied in the same message as independent verification. The completion evidence is a documented control and a passed exercise using synthetic data.
9.5 Reporting and third-party resilience
The Bundesbank's DORA guidance identifies BaFin as the competent recipient for relevant financial-entity incident reports and explains onward sharing with other authorities. It also describes the register of contractual ICT service arrangements. [S32]
For major ICT-related incidents, Article 5 of Delegated Regulation (EU) 2025/301 sets these clocks. [S88]
| Submission | Deadline |
|---|---|
| Initial notification | As early as possible; within 4 hours of major-incident classification and generally within 24 hours of awareness |
| Late classification | If classified as major after 24 hours of awareness, notify within 4 hours of classification |
| Intermediate report | Within 72 hours of initial notification, even without changes; update without undue delay and upon recovery of regular activities |
| Final report | Within 1 month of the intermediate report or latest updated intermediate report |
Article 5 allows submission by noon of the next working day when a deadline falls on a weekend or bank holiday. Credit institutions, central counterparties, trading-venue operators and NIS2 essential or important entities cannot use that extension for initial or intermediate reports. Authorities may also exclude other significant or systemic financial entities. If unable to meet a deadline, inform the authority without undue delay, at the latest by that deadline, and explain why. [S88]
The practical requirement is to connect a service interruption to the organisation's classification and reporting process. A supplier may have an incident without every affected customer meeting the same reporting trigger. Record the determination and the information still required.
The register should be usable in an incident. A team should be able to identify the provider, critical function, technical owner, escalation contact and alternative arrangement. Testing that lookup is more useful than treating an untested spreadsheet as completed resilience.
9.6 Validation and closure
Use the observed financial-domain signals for owned-account checks. Confirm employment, service ownership, account status and earlier remediation before containment. Related services can belong to the same legal entity. A current advertisement does not establish that institution's internal access was obtained.
The narrow payment-card observations do not measure national card exposure. Keep customer-fraud controls and staff-account response under separate accountable owners.
Key takeaways
- 01
Validate financial-domain observations against the relevant accounts and services.
- 02
Historical document disclosure and current access require different evidence.
- 03
Validate identities, payment changes and provider dependencies through established bank processes.
Chapter 10
Regulatory and national response
This chapter answers
Which obligations should a German incident team operationalise before an event?
In short
Germany's NIS2 implementation is in force. In-scope entities must connect security measures, management oversight, registration and significant-incident reporting. Personal-data reporting and financial-sector reporting have different triggers and recipients. An incident may require several assessments. Name the decision-makers, preserve awareness times and rehearse a joint exercise with a critical provider.
This chapter is a planning aid, not legal advice.
10.1 Rules and responsibilities
The table records selected operational obligations. It does not determine whether a particular candidate organisation is in scope or reproduce every statutory exception.
| Rule | Applies where | Operational requirement | Decision owner |
|---|---|---|---|
| BSIG §30 | Entity qualifies under the relevant German scope provisions | Proportionate security risk measures, including continuity, supplier security and authentication; document compliance [S26] | CISO and management |
| BSIG §32 | In-scope entity has a significant incident | Early warning and notification without undue delay, at the latest within 24 and 72 hours of awareness, respectively. Final report normally within one month after notification [S27] | Incident lead and legal |
| BSIG §33 | Relevant entity first qualifies or qualifies again | Register within three months; update changed details within the statutory period [S28] | Legal and accountable entity owner |
| BSIG §38 | Management of an in-scope entity | Implement and oversee required risk measures; attend recurring training [S29] | Management |
| GDPR Articles 33–34 | A personal-data breach meets the relevant risk test | Controller notifies the competent authority without undue delay and, where feasible, within 72 hours; assess affected-person communication [S31] | Data-protection officer and controller |
| DORA | Relevant financial entity and major ICT-related incident | Apply the notification sequence and exceptions in §9.5; use the competent-authority route and maintain ICT third-party records [S32][S88] | Financial incident function and compliance |
Germany's implementation commenced on 6 December 2025. Registration timing depends on when the entity qualifies; it is not a recurring annual grace period. Scope must be established for the actual legal entity and activity. [S30][S28]
A separate parliamentary announcement records passage of a cybersecurity bill on 12 November. Passage alone does not establish commencement. It does not replace the in-force obligations used in this table without checking the promulgated law and applicable date. This report takes no position on the policy choices. [S85]
10.2 Notification clocks require an awareness record
For the relevant BSIG duty, the early and subsequent notices have different content and deadlines. The final-report timing follows the incident notification. Where an incident continues, the statute provides for a progress report and subsequent final reporting. [S27]
Under GDPR, the controller's authority notification depends on the risk to people. A processor informs the controller without undue delay. High risk can require communication to affected people, subject to the applicable exceptions. Document the assessment even where the conclusion is not to notify. [S31]
Do not start every clock from a convenient press-release date. Preserve when the relevant entity became aware, what was known then and which information arrived later. Legal should record the trigger assessment and the rationale for any staged or updated submission.
10.3 One incident, several competent routes
The recipient can depend on the entity, sector and jurisdiction. A personal-data question should identify the competent supervisory authority. A financial incident follows the relevant financial framework. A police report addresses a different purpose from regulatory notification.
The incident record should identify each route, named decision-maker and deadline. A shared chronology can support several notifications, but a submission under one regime should not be assumed to satisfy another without checking the applicable rule.
Supplier escalation should support those determinations. Ask for the discovery and awareness timeline, affected service, data categories, possible customer impact and current containment status. Separate confirmed facts from estimates so that legal can decide with the evidence available.
10.4 Management oversight needs operational evidence
BSIG management duties concern implementation and oversight, not merely receipt of a security presentation. The statutory training duty also requires recurring attention to risk and the consequences for the entity's services. [S29]
For board reporting, connect each material dependency to an accountable owner and a test. Examples include identity recovery, a critical supplier's incident notice, a service restoration and an operational fallback. Record the remaining exception, its owner and the date for closure.
The report does not offer a universal fine figure or infer a liability outcome from an incident. Those depend on the applicable provision, entity and facts. Verify the current legal text before relying on an enforcement conclusion.
10.5 National response and practical cooperation
The BKA/ZIT and state-police notices demonstrate investigations across malware, DDoS and fraud infrastructure. Their publication does not certify that every affected system is clean. Relevant notifications should enter the organisation's own response and validation process. [S23][S24][S25][S43]
The Berlin authority's notification statistics also show why breach management cannot be reduced to malware counts. The population includes personal-data incidents under a different reporting definition. It should inform notification readiness without being added to national police or listing totals. [S35]
10.6 A reporting rehearsal
| Exercise prompt | Required decision | Evidence |
|---|---|---|
| Provider reports possible data access | Does the controller have sufficient awareness to start assessment? | Awareness time and legal rationale |
| Critical service becomes unavailable | Does the sector significance test apply? | Service-impact record and classification |
| Scope remains uncertain | What can be submitted and updated now? | Staged notice and missing-information list |
| People face likely high risk | Is direct communication required and accessible? | Approved notice and support route |
| Incident continues | Which progress or final report is due? | Reporting calendar with responsible owner |
Run the exercise with security, legal, data protection, communications and the provider. The pass condition is a documented decision at the required time, including the uncertainty. It is not an unrealistically complete forensic report produced immediately.
Key takeaways
- 01
Germany's NIS2 implementation is in force; establish entity-specific scope now.
- 02
Preserve awareness times and assess each reporting framework separately.
- 03
Management oversight should be demonstrated through owned controls, tested decisions and recorded exceptions.
Chapter 11
Outlook and conditional planning
This chapter answers
Which preparations follow from the evidence, and what would change the assessment?
In short
Prepare for extortion, identity misuse, supplier-held data loss and service disruption. These are supported response priorities, not probabilities of a future attack. The listing population provides comparative visibility; the public cases provide mechanisms and consequences. The selected domain observations support validation, with no sector ratings. Assign an owner to the signals below and revise the plan when a relevant signal changes.
11.1 Key judgements
| Judgement | Confidence | Basis and limit |
|---|---|---|
| Extortion readiness remains a cross-sector priority | High for observed listing visibility | 457 German listings across 70 labels; listings are claims [S2] |
| Leading claimant labels do not cover most observed listings | High for the retained population | Top three hold 161; the other 296 carry different labels [S2] |
| Identity response should include sessions and privileged grants | High as a defensive implication | Campus disclosure supports identity-focused containment; not a national prevalence estimate [S11] |
| Customer protection can continue after technical containment | High for documented follow-on patterns | Official patient and booking warnings; no estimate of total losses [S16][S18] |
| Industrial continuity needs physical verification | High as a planning requirement | Official cooling warning; no common operator inferred [S21] |
| Domain exposure supports account validation | High for the selected portfolio | Dated device, credential and listing observations; no sector ratings or prevalence claim [S87] |
Confidence attaches to each proposition. It is not an assessment of an actor's overall capability. None of the rows predicts an attack date, a national loss amount or a particular organisation's likelihood of being attacked.
11.2 Signals that should change the response
| Current boundary | New evidence | Changed decision |
|---|---|---|
| Advertisement only | Validated current account or access path | Activate identity or incident containment |
| Availability incident | Confirmed data access | Add data scoping and notification assessment |
| Possible industrial access | Independent physical evidence of manipulation | Escalate to the operational safety process |
| Supplier incident scope uncertain | Reliable affected-data and customer list | Target support and legal communications |
| Attribution unresolved | A sourced official or technical attribution | Update actor assessment without rewriting the earlier evidence |
| Changed domain exposure | Validated and dated account or service observations | Reconcile scope and ownership before changing response priorities |
11.3 Conditional scenarios
Scenario 1: a current identity is advertised. If validation confirms an active account tied to a critical service, review its privilege, sessions and recent use. The signal is validation against owned systems. The decision is proportionate containment, not a blanket reset based solely on an advertisement.
Scenario 2: a provider reports customer-data access. If the provider supplies credible evidence of affected categories, involve legal and customer operations immediately. The signal is a scoped provider notice. The decision is a timed notification assessment and support plan while technical investigation continues.
Scenario 3: a claimant announces disruption. If monitoring confirms degraded service, activate the continuity and mitigation owners. The signal is the essential task failing, not only the actor's message. The decision is a bounded public status statement and the agreed fallback.
Scenario 4: industrial readings become inconsistent. If an independent measurement shows an unsafe condition, use the approved operational procedure. The signal is physical evidence and the operator's assessment. The decision includes stock protection and safety, rather than an IT-only restoration objective.
Scenario 5: official infection notification arrives. If a trusted authority identifies an owned asset, validate its status and related identities. The signal is a matched asset and supporting evidence. The decision is host remediation, session revocation and a documented investigation, even after infrastructure disruption.
These are planning assumptions. They are not measured response-time intervals, forecasts that an event will occur, or claims that any organisation has implemented the recommended controls.
11.4 What this report cannot project
The observed listing increase over a full-year baseline does not establish a future growth rate. Underground post volume cannot supply an attack probability. Incomplete organisation measurement cannot support a national security score.
Public cases also reflect disclosure choices. A sector with more detailed notices can appear more prominent without having a higher underlying incident rate. Use the cases to test response capabilities and the listing counts to understand visibility, while retaining those limits.
Key takeaways
- 01
Prepare for mechanisms and consequences supported by the evidence.
- 02
Assign a signal and decision owner to each scenario.
- 03
Change the assessment when new evidence resolves an uncertainty, rather than when an actor's rhetoric changes.
Chapter 12
Action plan
This chapter answers
What should teams deliver in the next 90 days, and how will leaders verify it?
In short
Start with alert ownership, account validation, critical-service continuity and reporting scope. Then test supplier notice, identity recovery, customer verification and operational access. By day 90, demonstrate restoration and a combined incident exercise to leadership. These are planning targets. A current material incident should enter the response process immediately, with no wait for the relevant planning period.
12.1 The 30/60/90-day plan
The actions are conditional on the organisation's services and legal scope. Use synthetic data and approved procedures for exercises.
1Days 1–30
Establish a route for identity, supplier and relevant public-claim alerts
- Owner
- CISO and threat intelligence
- Completion evidence
- Named alert source, owner and escalation test
Validate each relevant identity alert for ownership, freshness, status and privilege
- Owner
- Identity and SOC
- Completion evidence
- Dated disposition for every received alert
Confirm session revocation and privileged-access review for validated incidents
- Owner
- Identity and application teams
- Completion evidence
- Revocation evidence and grant review
Identify critical services and their minimum acceptable fallback
- Owner
- Business continuity and service owners
- Completion evidence
- Service map and approved fallback
Establish BSIG, GDPR and sector-reporting scope and awareness recording
- Owner
- Legal, data protection and CISO
- Completion evidence
- Obligation register and clock exercise
Approve status wording for claims, outages and possible data access
- Owner
- Communications and incident lead
- Completion evidence
- Templates with update owner and timing
2Days 31–60
Test incident notice and evidence exchange with a critical provider
- Owner
- Supplier management and legal
- Completion evidence
- Timed exercise and unresolved items
Test account recovery without relying solely on ordinary email
- Owner
- Identity and service desk
- Completion evidence
- Synthetic recovery case and audit trail
Review administrative grants, delegated mail access and recovery methods
- Owner
- Identity and application owners
- Completion evidence
- Reconciled privilege inventory
Review internet-facing remote support and verify applicable updates
- Owner
- Infrastructure
- Completion evidence
- Version, patch and investigation decisions
Test application authorisation on critical internal functions
- Owner
- Application security
- Completion evidence
- Approved test record and fixes
Inventory legitimate mail senders and plan controlled DMARC enforcement
- Owner
- Messaging
- Completion evidence
- Validated sender inventory and change plan
Exercise contracted mitigation and customer-service fallbacks
- Owner
- Infrastructure and service owners
- Completion evidence
- Minimum user task works during test
Restrict industrial maintenance access under approved procedures
- Owner
- OT, facilities and supplier owner
- Completion evidence
- Access inventory and revocation test
Verify independent operational alarms and fallback storage where relevant
- Owner
- Operations and safety
- Completion evidence
- Witnessed alarm and contingency exercise
Publish trusted payment and contact verification routes
- Owner
- Customer operations and fraud
- Completion evidence
- Public guidance and staff exercise
3Days 61–90
Restore a critical service in an isolated environment
- Owner
- Infrastructure and business continuity
- Completion evidence
- Restore record against agreed objective
Exercise supplier data loss, claim handling and reporting together
- Owner
- Incident lead, legal and communications
- Completion evidence
- Timed decisions and staged notices
Review retention and access on sensitive shared document stores
- Owner
- Data owners and data protection
- Completion evidence
- Retention decisions and access evidence
Present results and owned exceptions to management
- Owner
- CISO and management
- Completion evidence
- Scorecard with owners and closure dates
12.2 What each role should own
Board and management. Ask which essential function fails first, which provider can trigger that failure and what evidence shows the fallback works. Require a named owner and closure date for each exception. Do not substitute a national threat headline for local completion evidence.
CISO and incident lead. Preserve the evidence type on every claim. Connect security findings to a service owner, legal determination and customer impact assessment. Escalate uncertainty explicitly when the available evidence cannot support a decisive statement.
SOC and identity. Validate relevant accounts and assets, review sessions and grants, and retain the decision trail. Reconcile the record with known remediation so that old exposure is not repeatedly treated as a new event.
Infrastructure and application teams. Verify installed versions and effective permissions, then investigate earlier exposure where appropriate. Demonstrate service restoration through the user's task. Keep recovery dependencies outside the same failure path where the architecture permits it.
OT and facilities. Own physical readings, safe-state decisions and operational fallback. Security supports the access investigation; it should not replace the equipment owner's safety judgement with a generic instruction to disconnect everything.
Fraud and customer operations. Prepare independent verification of messages and payment requests. Explain the limits of deletion claims and avoid asking affected people to disclose additional sensitive material through an unverified channel.
Legal, data protection and communications. Record awareness, risk and reporting determinations. Make public statements specific to what is known. Preserve qualifications when translating or summarising the incident for another audience.
12.3 Leadership scorecard
These are recommended planning targets, not measured control performance in German organisations.
| Measure | Target | Evidence |
|---|---|---|
| Alert ownership | In place by day 30 | Successful routing test |
| Relevant received alerts | Every alert has a dated disposition by day 30 or an owned exception | Case register |
| Reporting scope | Recorded by day 30 | Entity-specific obligation register |
| Critical-provider exercise | Completed by day 60 | Timed notice and evidence exchange |
| Privileged recovery | Tested by day 60 | Synthetic case and session checks |
| Minimum service | Demonstrated by day 60 | User-journey and fallback record |
| Restoration | Passed or dated remediation by day 90 | Isolated restore test |
| Management review | Completed by day 90 | Signed decisions and exception owners |
12.4 How Brandefense can help
Brandefense combines threat intelligence, digital risk protection and external attack-surface management. Relevant use cases include validating identity exposure, monitoring impersonation and tracking external findings through a disposition process. [S44]
Request a demonstration using a representative owned domain and a critical service. Ask how an alert reaches your team, which evidence supports it and what demonstrates closure. A product demonstration does not establish that your organisation has been assessed in this report.
Key takeaways
- 01
Establish evidence ownership and reporting scope by day 30.
- 02
Test provider, identity, customer and operational response by day 60.
- 03
Demonstrate restoration and report owned exceptions to management by day 90.
Chapter A
Appendix A. Scope and evidence boundaries
This chapter answers
Which populations, dates and limits apply to the report's findings?
Comparison countries were selected as examples to provide context only; they do not imply a ranking or political assessment.
A.1 Window and dates
The requested observation window is 1 January–22 November 2026. Underground observations use post publication time. Ransomware observations use listing publication dates where available. Public reporting distinguishes incident, disclosure and update dates where the source provides them.
Research was conducted on 22 November local time. The cut-off day was still in progress during collection. The report therefore does not claim a completed end-of-day census. The recent-activity section covers information available for 8–22 November and includes clearly labelled updates and enforcement announcements.
BSI's annual assessment covers an earlier July-to-June period. Police statistics and data-protection notifications retain their own years and definitions. None is added to the report's listing population. [S3][S4][S35]
A.2 Sources and priority
Own-source populations are underground material from Odin by Brandefense and ransomware listings from Ransompedia. External primary sources include BSI, BfV-related warnings, the Bundestag, police, prosecutors, data-protection authorities and organisational disclosures. Wider reporting supports naming and provides context.
The five-country tables use Ransompedia listing counts. The observed Germany-associated post total and platform/language observations describe publication visibility. Broader content coverage remains incomplete. The dated domain exposure observations retain their own assessed denominator. The distinction is reflected in the relevant chapters rather than hidden in a general disclaimer. [S1][S2]
A.3 Ransomware listings
Ransompedia records 457 German listings within the requested window. The comparator counts are France 295, Netherlands 77, Poland 56 and Austria 61. Listings record criminal claims. They do not establish unique intrusions, encryption, copied data, payment or affected people. [S2]
Prior-year figures describe all of 2025. The current-window and prior-year figures are not matched periods. Comparisons between countries also reflect different economic scale, disclosure and visibility. They are not national risk rankings.
The group-label and monthly tables reconcile to the same German listing total. Industry information describes the listing population. It does not measure sector prevalence. Smaller listing-sector groups are combined in the public chart data.
A.4 Organisation exposure
The snapshot dated 21–22 November 2026 covers 249 of 250 selected domains. It records at least 636,397 device associations, 33,915 employee-device associations and 125,422,358 credential-record associations. In all, 107 observed domains had a credential listed for sale. These are overlapping records, not unique people or confirmed intrusions. [S87]
Legal ownership and independent organisation identity remain unvalidated. Related domains can represent the same operator. One industry-domain record is unavailable. The national Atlas headline for the observed portfolio is 61 out of 100. No sector or organisation ratings are published.
Passive hygiene results remain unavailable as a completed dataset. An unfinished or unobserved measurement is not an absent control. Device, credential, listing, card and incident populations remain separate.
A.5 Card material
The figures describe payment-card records in the retained post corpus. They do not measure national card exposure, card theft or fraud loss.
No personal card data or issuer-level identifier is published. The card population must not be added to credentials, devices or incident counts. The result and its availability are stated in §8.5.
A.6 Incident coverage
The incident register contains documented cases, campaign notices, enforcement, exclusions and unresolved discoveries. Those categories are not a single incident denominator. The report does not claim exhaustive national incident coverage.
Some source pages and late discoveries remain unresolved. Some incidents are never disclosed. A supplier event can affect several organisations, and several updates can concern one event. Absence from this report does not establish absence of an incident.
A.7 Evidence and confidence
| Label | Meaning |
|---|---|
| Observation | A retained record supports the stated measurement or publication |
| Claim | A source asserts access, impact or possession that is not independently established here |
| Official confirmation | The organisation or authority confirms the particular proposition stated |
| Analyst inference | A stated interpretation with a basis and a limit |
| Analytical model | A planning relationship, without a proven shared operator or transaction chain |
High confidence attaches to a reproducible quantity or well-supported proposition. It does not certify an actor's entire assertion. Unverified does not mean fabricated, and a confirmed incident does not confirm every advertised detail.
A.8 Naming, privacy and neutrality
Organisations are named as incident subjects only where official confirmation and wider reporting support the specific case. Other subjects use broad roles. Some source titles and article addresses are withheld to preserve those roles; the supporting passages remain in the private evidence register.
Actors' motives and state associations remain attributed to their sources. The report takes no position on political parties, elections, governments, territory, religion or foreign-policy decisions. No personal identities, criminal contacts, victim access details or sensitive sample records are published.
A.9 Verification limits
Every external proposition used was checked against a supporting passage. The legal table distinguishes commencement, scope and reporting triggers. The vulnerability reference was checked against NVD and the vendor's remediation guidance. It is not evidence of a German exploitation incident. [S33][S34]
Automatic consistency checks do not establish exhaustive coverage or release approval. The English master and German translation retain the same figures, source identifiers, case identifiers and directives. The English edition prevails if the language editions differ.
Appendix B. Evidence register
The case register below identifies what each case establishes. It is not a list of unique national cyberattacks. Card metadata and in-text rows must retain the same limitations.
| Case | Date or phase | Sector | Evidence and boundary |
|---|---|---|---|
| A | August; later notices | Public administration | Data copying confirmed; individual impact review continues |
| B | 19 November | Public property | Security incident; possible outflow unresolved |
| C | 6–9 November | Education | Credentials and campus access; no whole-population theft total |
| D | 10 November | Education | Employee records; students and passwords excluded in that notice |
| E | 10–11 November | Education | Separate webmail event; possible credential capture |
| F | 19 September | Education | Registration records; no public-leak conclusion |
| G | February–March | Healthcare | Intrusion and later patient extortion, one case |
| H | 22 September | Medical technology | Limited internal systems; stated operational exclusions |
| I | 18 November warning | Industrial health supply | Cooling attacks reported; no German blood-bank impact established |
| J | May | Supplier services | Possible child-data exposure; no verified count |
| K | July notice | Hospitality | Supplier-held records and later fraudulent messages |
| L | February | Transport | Booking and information availability disruption |
| M | 17 November | Municipal administration | Unauthorised access; theft still investigated |
| N | 18–20 November | Communications | Attack and service restoration; no theft established |
| O | 15–18 November | Healthcare | One hospital event and recovery phases |
| P | 13–20 November | Hosting | Infrastructure access; later higher-risk instance update |
| Q | 12 November notice | Social services | Partial IT restrictions; in-person work continues |
| R | September–November | Political organisation | Earlier access, recent notice; no verified people total |
| S | 8 November post | Banking-related services | Administrator-tenant offer; live access and domestic legal entity unverified |
| T | 16 November post | Cross-sector documents | German-organisation file offer; origin, age and possession unverified |
| U | 2 February post | Project services | Contact/budget database offer; authenticity and acquisition unverified |
| V | 6 October post | Cross-sector documents | Administrative/financial release assertion; files and current access unverified |
| W | 17 February posts | Automotive | Documents and exploit offered; vulnerability and possession unverified; repeat is not another incident |
| X | 5 January post | Identity data | Buyer demand for identities and addresses; no possession evidence |
| Y | 8–17 November | Hospitality supplier | Guest-data access, fraudulent messages and rebuilt-service restart; one supplier incident |
| Z | 8–20 November | Energy services | IT/customer-service disruption; energy supply maintained |
B.1 Reading cross-case relationships
An affected institution can appear in more than one case when separate notices describe different events. Conversely, several dates can be phases of a single case. The Tübingen notices are separate; the hospital intrusion and subsequent patient messages are one case.
No cross-case common operator is asserted. Named ransomware labels belong to the listing population, and official campaign warnings retain their geographic limits. A public disclosure is not automatically matched to a criminal listing unless the report explicitly establishes that relationship.
B.2 Post quotations
| Card | Case | Post date | Proposition |
|---|---|---|---|
| B01 | S | 2026-11-08 | An advertisement asserts administrator access to a banking-related CRM tenant and describes customer information. |
| B02 | T | 2026-11-16 | An advertisement offers files said to come from German organisations. |
| B03 | U | 2026-02-02 | A seller describes project-contact and budget records associated with a German portal. |
| B04 | V | 2026-10-06 | A post asserts release of administrative, financial and corporate email material from an organisation described as German. |
| B05 | W | 2026-02-17 | A seller advertises automotive documents and separately an IDOR exploit. |
| B06 | X | 2026-01-05 | A buyer requests German identity information and emphasises real home addresses. |
Appendix C. Sources
Every entry records the supporting-source review date. Some public addresses are shortened to protect role-based subjects. Access limitations are recorded separately; an indexed passage is distinguished from a successful page retrieval.
- S1Odin by Brandefense. Retained Germany underground post evidence. Own-source evidence; no public address. Verified on 22 November 2026.
- S2Ransompedia. Germany and regional-peer ransomware listing aggregates. Own-source evidence; no public address. Verified on 22 November 2026.
- S3BSI. Die Lage der IT-Sicherheit in Deutschland 2026. https://bsi-lagebericht.de/ Verified on 22 November 2026.
- S4BKA. Bundeslagebild Cybercrime 2025, press statement, 12 May 2026. https://www.presseportal.de/blaulicht/pm/7/6273717 Verified on 22 November 2026.
- S5BSI. Vorfall: Phishing-Angriffe via Messenger. https://bsi-lagebericht.de/node/44 Verified on 22 November 2026.
- S6Deutscher Bundestag. Signal-Phishing-Angriffe beschäftigen Digitalausschuss, 7 May 2026. https://www.bundestag.de/presse/hib/kurzmeldungen-1174610 Verified on 22 November 2026.
- S7Deutsche Bahn. Interim Report 2026: Digitalization. https://zbir.deutschebahn.com/2026/en/interim-group-management-report-unaudited/digitalization-and-technology/digitalization/ Verified on 22 November 2026.
- S8Tagesschau. Booking systems disrupted by cyberattack, 18 February 2026. https://www.tagesschau.de/inland/gesellschaft/bahn-stoerung-it-cyberangriff-100.html Verified on 22 November 2026.
- S9Berlin data-protection authority. Cyberangriff auf das Landesnetz Berlin. https://www.datenschutz-berlin.de/datenschutz/hinweise-zum-hackerangriff-auf-berlin/ Verified on 22 November 2026.
- S10Berlin Senate. Cyberangriff auf das Landesnetz: Informationen für Bürgerinnen und Bürger. https://www.berlin.de/cyberangriff/informationen-fuer-buergerinnen-und-buerger/ Verified on 22 November 2026.
- S11Technische Universität München. IT-Sicherheitsvorfall an der TUM, 6 November; update 9 November 2026. https://www.tum.de/aktuelles/alle-meldungen/pressemitteilungen/details/it-sicherheitsvorfall-an-der-tum Verified on 22 November 2026.
- S12BR24. Erneut Hackerangriff auf bayerische Uni: TUM betroffen. https://www.br.de/nachrichten/bayern/erneut-hackerangriff-auf-bayerische-uni-tum-betroffen,VaCjyBT Verified on 22 November 2026.
- S13Universität Tübingen. IT-Sicherheitsvorfall an der Universität Tübingen, 10 November 2026. https://uni-tuebingen.de/universitaet/aktuelles-und-publikationen/pressemitteilung/it-sicherheitsvorfall-an-der-universitaet-tuebingen/ Verified on 22 November 2026.
- S14Universität Tübingen. Nutzersperrungen nach Angriff auf Webmail, 11 November 2026. https://uni-tuebingen.de/newsfullview-landingpage/article/nutzersperrungen-nach-angriff-auf-webmail/ Verified on 22 November 2026.
- S15SWR / Tagesschau. Cyberangriff auf die Uni Tübingen: 3.900 Datensätze abgegriffen. https://www.tagesschau.de/inland/regional/badenwuerttemberg/swr-cyberangriff-auf-die-uni-tuebingen-3-900-datensaetze-abgegriffen-100.html Verified on 22 November 2026.
- S16LfDI Mecklenburg-Vorpommern and LKA Mecklenburg-Vorpommern. Hospital incident and follow-on patient extortion warning, 26 March 2026. https://www.datenschutz-mv.de/ (article URL withheld) Verified on 22 November 2026.
- S17LfDI Mecklenburg-Vorpommern. Photo-service supplier incident and possible child-data exposure, 22 May 2026. https://www.datenschutz-mv.de/ (article URL withheld) Verified on 22 November 2026.
- S18LfDI Mecklenburg-Vorpommern. Hotel-software incident and follow-up fraud warning, 29 July 2026. https://www.datenschutz-mv.de/ (article URL withheld) Verified on 22 November 2026.
- S19Fresenius Medical Care. Statement Regarding Cybersecurity Incident, 22 September 2026. https://freseniusmedicalcare.com/en/media/newsroom/statement-regarding-cybersecurity-incident/ Verified on 22 November 2026.
- S20Hessenschau. Hacker greifen Fresenius Medical Care an, 23 September 2026. https://www.hessenschau.de/wirtschaft/hacker-greifen-fresenius-medical-care-an-mehrere-systeme-betroffen-v2,fresenius-hackerangriff-100.html Verified on 22 November 2026.
- S21LKA Nordrhein-Westfalen. Cyberangriffe auf Kühlanlagen, 18 November 2026. https://www.presseportal.de/blaulicht/pm/58451/6392866 Verified on 22 November 2026.
- S22State ministry. IT security incident at a state property agency, 19 November 2026. https://mkbwk.rlp.de/ (article URL withheld) Verified on 22 November 2026.
- S23BKA and ZIT. Operation Endgame action, 12 November 2026. https://www.presseportal.de/blaulicht/pm/7/6389442 Verified on 22 November 2026.
- S24BKA and ZAC NRW. Action against botnet infrastructure, 20 March 2026. https://www.presseportal.de/blaulicht/pm/7/6239788 Verified on 22 November 2026.
- S25LKA Schleswig-Holstein. DDoS and attempted-extortion investigation, 31 August 2026. https://www.presseportal.de/blaulicht/pm/2256/6343014 Verified on 22 November 2026.
- S26Gesetze im Internet. BSIG 2025 §30: risk-management measures. https://www.gesetze-im-internet.de/bsig_2025/__30.html Verified on 22 November 2026.
- S27Bundesgesetzblatt / Bundesnetzagentur. NIS2 implementation act, BGBl. 2025 I Nr. 301, original §32 and Article 30. https://www.bundesnetzagentur.de/DE/Fachthemen/ElektrizitaetundGas/Versorgungssicherheit/IT_Sicherheit/_DL/NIS-2-Richtlinie.pdf?__blob=publicationFile&v=1 Verified on 22 November 2026.
- S28Gesetze im Internet. BSIG 2025 §33: registration. https://www.gesetze-im-internet.de/bsig_2025/__33.html Verified on 22 November 2026.
- S29Gesetze im Internet. BSIG 2025 §38: management duties. https://www.gesetze-im-internet.de/bsig_2025/__38.html Verified on 22 November 2026.
- S30Bundesregierung. NIS2 implementation commenced on 6 December 2025. https://www.bundesregierung.de/breg-de/service/fragen-und-anworten/nis-2-richtlinie-deutschland-2373174 Verified on 22 November 2026.
- S31BfDI. Info 1: GDPR and BDSG, Articles 33–34. https://www.bfdi.bund.de/SharedDocs/Downloads/DE/Broschueren/INFO1.pdf?__blob=publicationFile&v=24 Verified on 22 November 2026.
- S32Deutsche Bundesbank. DORA incident reporting and third-party information register. https://www.bundesbank.de/de/service/meldewesen/finanzaufsicht/europaeisches-meldewesen/digital-operational-resilience-act-dora--951046 Verified on 22 November 2026.
- S33NVD. CVE-2026-1731 record. https://nvd.nist.gov/vuln/detail/CVE-2026-1731 Verified on 22 November 2026.
- S34BeyondTrust. Security Advisory BT26-02. https://www.beyondtrust.com/trust-center/security-advisories/bt26-02 Verified on 22 November 2026.
- S35Berlin data-protection authority. Höchststand bei Datenschutzvorfällen, 22 October 2026. https://www.datenschutz-berlin.de/pressemitteilung/hoechststand-bei-datenschutzvorfaellen/ Verified on 22 November 2026.
- S36LfV Baden-Württemberg. APT28 greift Think Tanks in Europa an, 15 January 2026. https://www.verfassungsschutz-bw.de/,Lde_DE/Startseite/Arbeitsfelder/APT28+greift+Think+Tanks+in+Europa+an Verified on 22 November 2026.
- S37LfV Baden-Württemberg. Remote IT-worker warning, 24 July 2026. https://www.verfassungsschutz-bw.de/,Lde_DE/Startseite/Arbeitsfelder/Nordkoreanische+IT-Worker Verified on 22 November 2026.
- S38LfV Baden-Württemberg. Maritime-sector espionage in seven EU states, 25 September 2026. https://www.verfassungsschutz-bw.de/,Lde_DE/Startseite/Arbeitsfelder/Chinesische+Cyberspionage+gegen+den+maritimen+Sektor Verified on 22 November 2026.
- S39Deutsche Bank. Statement on recent media reports, 17 November 2026. https://www.db.com/news/detail/20261117-deutsche-bank-statement-on-recent-media-reports Verified on 22 November 2026.
- S40Euronews Russian edition. Reporting on historical bank-document disclosure, 18 November 2026. https://ru.euronews.com/2026/11/18/data-watchdog-investigates-massive-deutsche-bank-leak-involving-15-million-files Verified on 22 November 2026.
- S41CCZ Baden-Württemberg and LKA Baden-Württemberg. Email-account phishing investigation, 19 October 2026. https://generalstaatsanwaltschaft-karlsruhe.justiz-bw.de/ (article URL withheld) Verified on 22 November 2026.
- S42CCZ Baden-Württemberg. Fake-shop charge, 9 November 2026. https://generalstaatsanwaltschaft-karlsruhe.justiz-bw.de/pb/,Lde/Startseite/Presse/PM+v_+09_11_2026/?LISTPAGE=1222232 Verified on 22 November 2026.
- S43Bundesnetzagentur and partner authorities. Operation Herakles telephone-number action, 25 September 2026. https://www.bundesnetzagentur.de/SharedDocs/Pressemitteilungen/DE/2026/20260925_Herakles.html Verified on 22 November 2026.
- S44Brandefense. Platform and demonstration information. https://brandefense.io/request-a-demo/ Verified on 22 November 2026.
- S45BSI. Vorfall: Ausnutzung von Zero-Day-Schwachstellen. https://bsi-lagebericht.de/node/30 Verified on 22 November 2026.
- S47BKA and ZIT. Criminal marketplace operation, 8 May 2026. https://www.presseportal.de/blaulicht/pm/7/6271239 Passage verified on 22 November 2026. Direct retrieval: HTTP 200, full statement read on 22 November 2026.
- S48SWR. Webmail attack at Universität Tübingen, 12 November 2026. https://www.swr.de/swraktuell/baden-wuerttemberg/tuebingen/cyberangriff-auf-webmail-der-universitaet-tuebingen-100.html Verified on 22 November 2026.
- S49LMU Munich. Information about a personal data breach, 19 September 2026. https://www.lmu.de/en/about-lmu/structure/central-university-administration/media-relations-and-communications/press-room/press-release/information-about-a-personal-data-breach-f446c16e.html Verified on 22 November 2026.
- S50TechRadar. Defence supplier incident and company statement, 19 November 2026. https://www.techradar.com/ (article URL withheld) Verified on 22 November 2026.
- S51Police. Municipal unauthorised-access investigation, 17 November 2026. https://www.presseportal.de/blaulicht/pm/110975/6392627 Verified on 22 November 2026.
- S52Local utility communications provider. Cyber incident and service-restoration notices, 18–20 November 2026. https://www.stadtwerke-wittenberg.de/ (article URL withheld) Verified on 22 November 2026.
- S53County authority. Hospital incident and recovery notices, 15–18 November 2026. https://www.landkreis-greiz.de/ (article URL withheld) Verified on 22 November 2026.
- S54Hosting provider. Infrastructure-access incident, updates through 20 November 2026. https://help.contabo.com/ (article URL withheld) Verified on 22 November 2026.
- S55Municipal authority. Social-service provider communications notice, 12 November 2026. https://www.offenbach.de/ (article URL withheld) Verified on 22 November 2026.
- S56AfD. Public personal-data incident notification, 13 November 2026. https://www.afd.de/wp-content/uploads/2026/11/2026-11-13-_-Allgemeine-AfD-Info-an-Betroffene-_-afd-de-datenpanne.pdf Verified on 22 November 2026.
- S57Tagesschau. AfD confirms email-account attack, October 2026. https://www.tagesschau.de/inland/innenpolitik/afd-hackerangriff-102.html Verified on 22 November 2026.
- S58Reservation software supplier. Guest-data access and fraudulent messages; updates 12–17 November 2026. https://reservision.world/ (article URL withheld) Verified on 22 November 2026.
- S59Municipal energy utility. IT incident detected 8 November; updates through 20 November 2026. https://www.stadtwerke-achim.de/ (article URL withheld) Verified on 22 November 2026.
- S60State police and prosecution. Cultural-collection attack on 21 January; investigation notice 19 February 2026. https://www.medienservice.sachsen.de/medien/news/1095116 Verified on 22 November 2026.
- S61Professional chamber. Website malware, possible password interception and 23 January 2026 restoration. https://www.aknr.de/ (article URL withheld) Verified on 22 November 2026.
- S62Municipal authority. 16 April ransomware; 20 and 24 April 2026 recovery updates. https://www.sprendlingen-gensingen.de/ (article URL withheld) Verified on 22 November 2026.
- S63Municipal energy utility. 28 June incident and updates through 16 July 2026. https://gsw-kamen.de/ (article URL withheld) Verified on 22 November 2026.
- S64Municipal authority statement, republished by local newspaper. 16 July service-provider attack; 21 July 2026 statement. https://www.l-iz.de/ (article URL withheld) Verified on 22 November 2026.
- S65Research institute. File-server access; notification 25 June 2026. https://www.zois-berlin.de/ (article URL withheld) Verified on 22 November 2026.
- S66Research institute. July network-component incident; updates 17 and 20 August 2026. https://www.idos-research.de/ (article URL withheld) Verified on 22 November 2026.
- S67Memorial foundation. 8 October 2026 notice of renewed website attack, distinct from earlier August incident. https://www.sachsenhausen-sbg.de/ (article URL withheld) Verified on 22 November 2026.
- S68Municipal authority. 15 October 2026 cyberattack and recovery notice. https://www.eichstaett.de/ (article URL withheld) Verified on 22 November 2026.
- S69Hessian data-protection authority. Warning about accommodation-sector attacks and fraudulent guest messages; undated. https://datenschutz.hessen.de/datenschutz/it-und-datenschutz/in-hessen-haeufen-sich-weiterhin-cyberangriffe-im-zusammenhang-mit-beherbergungsbetrieben Verified on 22 November 2026.
- S70ENISA. Joint statement on cyber incidents in refrigeration systems across the EU, 4 November 2026. https://www.enisa.europa.eu/news/joint-statement-on-cyber-incidents-in-refrigeration-systems-across-the-eu Verified on 22 November 2026.
- S71State museum. Incident detected 19 June; personal-data update 31 July 2026. https://www.landesmuseum.de/ (article URL withheld) Verified on 22 November 2026.
- S72LfV Baden-Württemberg. Mach-O Man: Modulares macOS-Schadprogramm, 30 April 2026. https://www.verfassungsschutz-bw.de/,Lde_DE/Startseite/Arbeitsfelder/Macho-O-Man+Lazarus Verified on 22 November 2026.
- S73County authority. Preventive shared-system shutdown, 29 September 2026 notice. https://www.kreis-bad-duerkheim.de/ (article URL withheld) Verified on 22 November 2026.
- S75University hospital. External billing-provider incident notice, 22 May 2026; later final scope unresolved. https://www.uksh.de/ (article URL withheld) Verified on 22 November 2026.
- S76Enterprise software vendor. 29 April 2026 malicious development-package versions; vendor mitigation notice. https://community.sap.com/ (article URL withheld) Passage verified in the indexed primary notice on 22 November 2026. Direct article retrieval: HTTP 403.
- S77University. 23 February emergency maintenance; 25 February 2026 notice. https://www.uni-hamburg.de/ (article URL withheld) Verified on 22 November 2026.
- S79LfV Baden-Württemberg. Programmable-controller security warning, 20 August 2026. https://www.verfassungsschutz-bw.de/ (article URL withheld) Verified on 22 November 2026.
- S80Vehicle retailer. Service-provider security notice, published 16 October and updated 29 October 2026. https://www.jepsen-autogruppe.de/ (article URL withheld) Verified on 22 November 2026.
- S81BKA and ZIT. International investigation after online-banking fraud, 14 August 2026. https://www.presseportal.de/blaulicht/pm/7/6333329 Verified on 22 November 2026.
- S82Deutscher Bundestag. Response on possible federal interests affected by Berlin incident, 16 November 2026. https://www.bundestag.de/presse/hib/kurzmeldungen-1233476 Verified on 22 November 2026.
- S83University hospital. External billing-provider incident notice, 22 May 2026. https://www.ukbnewsroom.de/ (article URL withheld) Verified on 22 November 2026.
- S84Certification organisation. 27 October 2026 publication acknowledging September IT security incident. https://www.gut-cert.de/ (article URL withheld) Verified on 22 November 2026.
- S85Deutscher Bundestag. Cybersecurity bill passage announcement, 12 November 2026. https://www.bundestag.de/dokumente/textarchiv/2026/kw46-de-cybersicherheit-1231534 Verified on 22 November 2026.
- S86Defence technology supplier. Incident statement, 12 November 2026; detection at end of October. https://quantum-systems.com/ (article URL withheld) Verified on 22 November 2026.
- S87Brandefense. Blackhole device and credential associations and Atlas exposure observations for a selected German domain portfolio, snapshot 21–22 November 2026. Own-source evidence; no public address. Verified on 22 November 2026.
- S88European Commission. Delegated Regulation (EU) 2025/301, Article 5: major ICT-related incident reporting deadlines and exceptions. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32025R0301 Passage verified on 22 November 2026. Retrieval: full official text read.
- S89Polizei Rheinland-Pfalz. Zentrale Ansprechstelle Cybercrime (ZAC). https://www.polizei.rlp.de/die-polizei/dienststellen/landeskriminalamt-rheinland-pfalz/service-und-ansprechstellen/zentrale-ansprechstelle-cybercrime-zac Passage verified on 22 November 2026. Retrieval: full official page read.
- S90Have I Been Pwned. Frequently Asked Questions: stealer logs. https://haveibeenpwned.com/FAQs Verified on 22 November 2026.
- S91Mozilla. Mozilla Monitor: Frequently asked questions, combolist definition. https://support.mozilla.org/en-US/kb/mozilla-monitor-faq Verified on 22 November 2026.
- S92Salesforce Architects. Platform Multitenant Architecture: tenant environments. https://architect.salesforce.com/docs/architect/fundamentals/guide/platform-multitenant-architecture Verified on 22 November 2026.
- S93Salesforce. What Is CRM (Customer Relationship Management)? https://www.salesforce.com/crm/what-is-crm/ Verified on 22 November 2026.
Appendix D. Glossary
| Term | Plain meaning |
|---|---|
| Account recovery | The process for restoring legitimate access after loss or restriction |
| Awareness time | The recorded time at which the relevant entity knows enough to begin an incident assessment |
| BfV | Germany's federal domestic intelligence authority |
| BKA | Germany's federal criminal police authority |
| BSI | Germany's federal information-security authority |
| BSIG | The German statutory framework for the BSI and relevant information-security duties |
| CERT-Bund | The federal computer emergency response function within BSI |
| Combolist | A combined list of email addresses and passwords from different data breaches [S91] |
| Credential | Information used to authenticate, such as a password or token |
| CRM tenant | An organisation's separate environment within cloud customer relationship management software [S92][S93] |
| Data controller | The party determining why and how personal data are processed |
| Data processor | A party processing personal data for a controller |
| DDoS | Distributed traffic intended to make a service unavailable |
| DMARC | Email policy connecting authentication and sender-domain alignment |
| DORA | EU financial-sector digital operational resilience framework |
| Extortion listing | A criminal publication asserting a claim about an organisation |
| Identity session | An authenticated period of access that may need revocation separately from a password |
| LKA | A state criminal police authority |
| MFA | Authentication requiring more than one factor |
| NIS2 | EU directive implemented through national information-security duties |
| OT | Equipment and systems controlling or monitoring physical processes |
| Phishing | Deceptive contact intended to induce disclosure or an unsafe action |
| Ransompedia | Brandefense's ransomware-listing source; listings record claims |
| Stealer-log material | Records produced by information-stealing malware on infected devices, including login details and associated websites [S90] |
| Supplier dependency | A function, data store or access path controlled by another organisation |
| TLP:CLEAR | Information may be shared without restriction, subject to other applicable duties |
| ZAC (police) | Central police contact points for cybercrime affecting organisations [S89] |
| ZAC NRW | Zentral- und Ansprechstelle Cybercrime Nordrhein-Westfalen, the prosecution unit at Cologne's General Prosecutor's Office named in [S24] |
| ZIT / CCZ | Specialist prosecution structures named in the relevant official notices |
About this report
Prepared by Odin by Brandefense for security leadership, incident responders, fraud teams, legal and service owners. Version 1.0 uses the requested research window of 1 January–22 November 2026 and a separate recent-activity view for 8–22 November.
Domain ownership and incomplete evidence coverage remain explicit. The German edition is translated from the English master; the English edition prevails where the editions differ.
© 2026 Brandefense. All rights reserved.