AI
Summary: Researchers at Anthropic and EPFL released an August 10, 2026 preprint showing self-propagating payloads they call "mind viruses" can spread between AI agents through editable system prompt files like SOUL.md, which agents wrote the payload into in 88% of propagation attempts and infected the next agent 55% of the time. In a simulated six-agent coding collaboration and paired-agent chains modeled on OpenClaw, DeepSeek V3.2, Qwen 3.5 32B and Gemini 3 Flash adopted an AI supremacy payload while Claude Sonnet 4.6, GPT-5.4 and Claude Haiku 4.5 did not, and a review of roughly 2,000 Moltbook propagation attempts from about 400 authors found no successful agent-to-agent spread.
Why it matters: A one-paragraph warning in the system prompt cut spread to near zero, and 15 generations of adversarial optimization across more than 150 candidate payloads produced no strain that got past a single hop - the control exists, and the question is whether your agent harnesses carry it.
Read the original at The Hacker News ↗
- Topics
- AI Security
- Related
- SECURITY AFFAIRS AI-CYBERSECURITY NEWSLETTER ROUND 8 Cyber Risks · Nov 15, 2026Pentagon Renames CDAO ‘Super Intelligence Office’ Cyber Risks · Nov 15, 2026186 Fake AI Domains Redirect Users to Click Fix Lures, Browser Malware, and Search Hijackers Cyber Risks · Nov 12, 2026