LOTUSLITE V2 Thailand campaign analysis by Antares Research describes an assessment-themed download that led to a .NET backdoor. The researchers published their report on 15 November 2026 and focused on activity directed at Thailand. Their observed chain used a signed PDFgear application to load a malicious DLL, then set Windows persistence and contacted external infrastructure. Organizations handling regional policy or security assessments should review the delivery and host evidence before deciding whether their systems were affected.
Antares assesses with high confidence that LOTUSLITE V2 succeeds the earlier LOTUSLITE family. It associates that family with Mustang Panda, also called Earth Preta, using shared protocol habits, execution methods, persistence, and regional focus. That is the researchers’ attribution assessment, not proof that a named operator controlled every observed host. A cyber threat intelligence platform can help teams compare the reported behavior with their own telemetry.
LOTUSLITE V2 Thailand Campaign: What the Report Shows
The initial lure was a public file-sharing link offering an Indo-Pacific economic and security assessment. Antares says opening the link downloaded an archive with a similarly named launcher and a DLL. The launcher staged a trusted PDFgear executable beside the malicious library. When the signed application ran, it loaded that nearby DLL, making a trusted binary part of the execution chain.
This is an application-local loading problem, not a reported flaw in PDFgear itself. Antares observed two logon autostart routes and a changed PDF file association, both implemented through Windows registry modifications. The implant then established an outbound connection. Its updated protocol framing and application-layer encryption also mean detection built solely around older LOTUSLITE network signatures may miss this variant.
The DLL side-loading technique explains why a valid signature on the executable is insufficient reassurance. The suspicious relationship is between the signed program, the untrusted library placed beside it, and the persistence changes that follow. Antares reports this chain from its analysis; the report does not establish how many organizations received the lure or executed it.
What Should Defenders Check Now?
Begin with download and endpoint records for assessment-themed archives received through file-sharing services. On candidate Windows hosts, inspect whether a signed PDFgear process loaded an unexpected library from a newly staged location. Review nearby registry changes for logon autostart and PDF association modifications. Correlate those findings with outbound connections from the same process before escalating an incident.
- Scope the lureIdentify recipients and downloads of the assessment-themed archive.
- Check executionReview signed PDFgear launches and libraries loaded beside them.
- Inspect persistenceCompare logon autostart and PDF association changes with approved software activity.
- Correlate trafficReview outbound connections from the affected process and preserve related logs.
- Contain confirmed hostsIsolate affected systems and investigate the original delivery path.
These are our defensive checks derived from the behaviors Antares documented. A matching archive name alone is weak evidence, while a linked download, DLL load, registry change, and outbound connection form a stronger case. Preserve the original artifacts and timestamps before removing persistence. Use your incident response plan template to assign evidence, containment, and recovery owners.
How Certain Is the Attribution?
Antares gives high confidence to the successor relationship, citing protocol and host behavior shared with earlier LOTUSLITE. The report also connects that lineage and Thailand focus to Mustang Panda. Treat that actor connection as the researchers’ assessment and retain the observed behaviors as the more direct hunting basis. Host evidence should drive the local incident decision, regardless of who operated the campaign.
Brandefense Blackhole counted one dark-chat post matching LOTUSLITE from 18 August through 16 November 2026. The measured unit is a matching post across that 90-day period. It is a discussion signal, not evidence of a victim, infection, or active exploitation. The Antares report remains the basis for the campaign details and the checks above.
Frequently Asked Questions
Is LOTUSLITE V2 a confirmed successor to LOTUSLITE?
Antares Research assesses the successor relationship with high confidence from shared protocol and host behaviors. It remains a research assessment.
Does the report prove every Thai organization was targeted?
No. Antares describes a Thailand-focused campaign and an analyzed chain, but it does not provide a count of affected organizations.
What is the first useful host check?
Correlate a relevant archive download with a signed PDFgear process loading an unexpected nearby DLL, then review persistence and traffic.
