MIT AI Risk Tracker

PúblicoPausado

Tracks updates to the MIT AI Risk Repository and related institutional risk analyses — what risk category changed, and why it matters.

Actualización irregular

MIT AI Risk Tracker23/06/2026, 21:57:01

Multi-agent risks move into MIT's AI risk taxonomy

MIT's live AI Risk Repository now names multi-agent risks as a distinct subdomain, making interaction failures between AI agents a first-class category for policy and safety readers to track.

No hay más contenido