OpenAI’s Jev clone could help stop swarming agents

This development addresses risks from multiple AI agents acting in coordination, potentially mitigating misuse.

Auto-scored: Low risk as it is a mitigation-focused development; mitigation is a real but partial step.