A new preprint study finds that “mind viruses” can spread ideas and goals between AI agents, potentially redirecting agents toward harmful behaviours in multi-agent systems.
AI Mind Viruses Could Push Agents to Pursue Harmful Goals, Study Finds
Theoretical and practical perspectives on AI mind viruses, information dynamics, memetic propagation, and algorithmic safety.
A new preprint study finds that “mind viruses” can spread ideas and goals between AI agents, potentially redirecting agents toward harmful behaviours in multi-agent systems.