Skip to content
NewCybercrime Report 2026 · Eight Chapters · A Post Is Not An Incident → Explore The ChaptersCybercrime Report 2026 → Explore

Odin Research / What is / Malware and backdoors

What Is Aurora Stealer? The Go Infostealer Sold as a Service

A Golang stealer rented out to criminal teams since 2022. What it takes from a victim, how it reaches them and what its servers looked like in Odin Research data.

By
Odin by Brandefense
Updated
21 Nov 2026
Reading time
6 min read
Type
Infostealer

In short

Aurora Stealer is an information-stealing malware written in Go and sold as malware-as-a-service on Russian-speaking forums and Telegram since 2022. It steals browser data, cryptocurrency wallets, Telegram sessions and a screenshot, then sends them to its server. It can also download and run a second payload.

Key takeaways

What it is
A Golang infostealer first sold as a botnet in April 2022 and relaunched as a stealer in late August 2022, according to Sekoia.
How it works
It fingerprints the host with WMIC, collects browser, wallet and Telegram data, takes a screenshot, and sends base64-encoded JSON to its C2 over TCP.
Why it matters
Stolen browser cookies and saved logins let buyers walk into accounts without passwords or MFA prompts, and stealer logs are resold widely.
What Odin Research saw
18 Aurora and AuroraStealer endpoints on 9 hosts in Monograph 24, with 6 confirmed, including one German server running three C2 ports.
What to do first
Block cracked-software and fake wallet downloads, alert on the WMIC fingerprint sequence, and revoke sessions for any user whose machine ran a stealer.

What is Aurora Stealer?

Aurora Stealer is an information-stealing malware written in Go and rented to criminals as a service. It collects saved browser data, cryptocurrency wallets and Telegram sessions from Windows computers and sends them to its operator. It can also download a second malware payload. It has been sold since 2022.

Sekoia's Threat Detection and Research team described Aurora in November 2022. It first appeared in April 2022 as a Golang botnet advertised on Russian-speaking underground forums. After a quiet period, it was relaunched in late August 2022 as an infostealer on Telegram and forums. Sekoia tracked nine "traffers" teams that added Aurora to their toolsets between July and November 2022.

An infostealer grabs credentials and session data in one quick pass rather than staying for long-term control. Aurora is part of the same market as RedLine Stealer and Lumma Stealer, sold as malware-as-a-service.

How Aurora Stealer works

Sekoia's analysis describes a short, noisy sequence:

  1. Fingerprint. Aurora runs three WMIC commands to read the OS, GPU and CPU.
  2. Collect. It targets web browsers, cryptocurrency wallet browser extensions, desktop wallets and Telegram Desktop data, and takes a screenshot of the screen.
  3. Send. Stolen files are base64-encoded, wrapped in JSON and sent to the C2 server over TCP. The screenshot goes first. Sekoia saw port 8081 most often, and port 9865 as well.
  4. Load. Aurora can download a second-stage file to the temp folder under a random name and start it through PowerShell.

Because it is written in Go, a single binary carries everything it needs. That also makes it easy for an operator to run several listeners on one server, a pattern Odin Research saw.

The data it steals maps to ATT&CK techniques such as Credentials from Password Stores and Steal Web Session Cookie (T1539).

How attackers spread Aurora Stealer

Sekoia described two main delivery methods:

  • Fake wallet sites. Phishing pages that copy cryptocurrency wallet download pages and serve Aurora instead.
  • The "911 method". YouTube videos send viewers to fake "free software catalogue" sites, which host the payload on file-sharing services. The lure is usually cracked or free software.

The stolen data ends up as stealer logs. Buyers use the cookies and saved passwords for account access, fraud and wallet theft, or resell them to initial access brokers.

What Odin Research observed

Monograph 24 studied how stolen browser sessions are replayed past MFA. Aurora sits in its "steal" tier, beside StealC, RedLine and LaplasClipper.

  • 18 endpoints on 9 hosts. Odin counted 13 AuroraStealer endpoints on 7 hosts, 6 of them confirmed C2 servers, and 5 Aurora endpoints on 2 hosts, none confirmed.
  • A three-port hub. One AuroraStealer server on Hetzner in Germany exposed three confirmed C2 ports on 13 November 2026: 21, 443 and 8443. Odin links this pattern to the Go build, which makes multiple listeners easy.
  • Remote desktop certificates. Two other confirmed AuroraStealer endpoints, in the United States and Russia, answered on port 3389 with certificates named after Windows or virtual machine host names.
  • Suspected staging. The 5 Aurora endpoints were suspected staging on ports 443, 3000 and 8080, still seen on 17 and 18 November 2026.
  • A suspected mail server. One suspected AuroraStealer server in Russia exposed web and mail ports (443, 995, 110, 993, 143) and shared a TLS fingerprint with a suspected StealC server.

Monograph 24 publishes the confirmed AuroraStealer C2 addresses for blocking. The rest stay on watchlists.

How to detect Aurora Stealer

  • WMIC fingerprint. Sekoia uses the sequence of three WMIC queries for OS, GPU and CPU as a correlation rule. A user-folder binary running all three in a row is suspicious.
  • Browser vault access. Alert when a process that is not a browser reads Chromium cookie and login databases or wallet extension folders.
  • Outbound TCP to rare ports. Raw TCP uploads to ports such as 8081 or 9865 from workstations, shortly after a download, deserve review.
  • Exfiltration markers. Odin Research detects AuroraStealer uploads to a /aurora/gate path that carry BrowserCookies and CryptoWallets fields.
  • PowerShell from temp. A random-named file in the temp folder started by PowerShell right after a stealer run may be the second payload.

Monograph 24 carries a Suricata signature for AuroraStealer exfiltration and a Sigma rule for non-browser access to Chromium vaults. It adds a Sentinel KQL hunt joined with cookie-store activity and a YARA rule for Aurora and StealC vault manifests. All are in the Defense Kit.

How to respond to an Aurora Stealer infection

  1. Revoke sessions, not just passwords. Stolen cookies stay valid after a password change. Sign the user out everywhere and reset tokens.
  2. Reset every saved credential on the device, and move any cryptocurrency to new wallets.
  3. Check for the second stage. Aurora can drop another payload, so reimage the machine.
  4. Monitor for resale. Watch stealer log markets for the user's accounts. See How to respond to infostealer log exposure.

Aurora Stealer compared with other infostealers

StealerLanguageC2 styleMonograph 24 count
Aurora StealerGoBase64 JSON over TCP; multi-port servers18 endpoints, 6 confirmed
StealCNot covered hereHTTP panels with PHP gates7 endpoints, 2 confirmed
RedLine StealerSee profileSee our RedLine profile6 endpoints
Lumma StealerSee profileSee our Lumma profileNot in this study

Frequently asked questions about Aurora Stealer

What is Aurora Stealer?

Aurora Stealer is an infostealer written in Go and sold as malware-as-a-service. It steals browser data, cryptocurrency wallets, Telegram data and a screenshot, and can download a second payload.

Is Aurora Stealer still active?

Odin Research saw 18 Aurora and AuroraStealer endpoints in November 2026, 6 of them confirmed C2 servers. Activity levels change, so check current threat intelligence.

How does Aurora Stealer spread?

Sekoia reported fake cryptocurrency wallet download sites and the 911 method, where YouTube videos lead to fake free-software sites that host the payload on file-sharing services.

What does Aurora Stealer steal?

Saved data from web browsers, cryptocurrency wallet extensions and desktop wallets, Telegram Desktop data, system details and a screenshot of the victim's screen.

Who made Aurora Stealer?

Sekoia says it was advertised by a seller using the handle Cheshire on Russian-speaking underground forums. Many different buyers run their own campaigns with it.

How do I detect Aurora Stealer?

Watch for three WMIC queries for OS, GPU and CPU in a row, non-browser access to browser vaults and raw TCP uploads to rare ports. Odin Research publishes Suricata, Sigma, KQL and YARA rules in its Defense Kit.

What should I do if Aurora Stealer ran on my computer?

Reimage the device, sign out of all sessions, reset every saved password from a clean device and move any cryptocurrency to new wallets.

Aurora Stealer vs RedLine: what is the difference?

Both are infostealers sold as a service. Aurora is written in Go and sends base64 JSON over raw TCP. RedLine has its own builds and C2 design; see the RedLine Stealer profile for details.

Sources

  1. Sekoia, "Aurora: a rising stealer flying under the radar", 21 November 2022. www.sekoia.com/blog/aurora-a-rising-stealer-flying-under-the-radar (checked 21 November 2026)
  2. MITRE ATT&CK, T1539 Steal Web Session Cookie. attack.mitre.org/techniques/T1539/ (checked 21 November 2026)

Related tools and terms

All Odin Research explanations · Brandefense glossary

Odin Research

Get the rules, not just the definition

Every Odin Research report ships with verified indicators and detection rules you can load today.

Open the Defense Kit