LowNewsLLM-specific
Malware evasion made easier with AI — but behavior still gives it away
- Published
- Record updated
Summary
Anthropic reported that Russia-linked group Midnight Blizzard (GTG-20006) used Claude-driven agents that rebuilt its malware whenever security tools detected it. ReversingLabs countered that five versions of the same PowerShell stager, each with a different hash, kept identical behavior over 25 days, which left them open to detection. Analysts say AI makes rebuilding faster but does not change the malware's core malicious behavior.
Related items
- InfoBank of America spent millions on anti-money-laundering tech, but not enough to avoid a $1.2B fineSame vendor · CSO Online
- InfoSpaceX policy manager set to testify as part of NYC Council AI safety investigationSame vendor · CNBC Technology
- InfoDesigning more efficient quantum circuits with ClaudeSame vendor · Anthropic Research Blog
- InfoAnthropic taps nonprofits METR and SecureBio to help assess risks and safety practicesSame vendor · CNBC Technology
- MediumGHSA-84m7-p3x7-pcfv: MCP Python SDK: Unbounded Session Accumulation in StreamableHTTP Transport Enables Resource ExhaustionSame vendor · GitHub Advisory Database