Transparent Tribe Launches "Vibeware" — First Named Nation-State AI Malware Campaign

Bitdefender documented Pakistan-aligned APT36 using LLMs to mass-produce malware in 7 languages (Nim, Zig, Crystal, Rust, Go, C#, .NET). They call the strategy "Distributed Denial of Detection" (DDoD) — flooding targets with disposable polyglot binaries to overwhelm signature-based detection. Malware families use Slack, Discord, Supabase, and Google Sheets as C2. This is fundamentally different from CyberStrikeAI: that was AI-augmented attacks with existing tools; this is AI-generated attack tools at scale. What to do: Your EDR must shift to behavioral process monitoring. Signature matching is now obsolete for this threat class. Monitor for outbound connections to Slack/Discord/Supabase from unsigned binaries. Bitdefender | The Hacker News