Incidents Models AI Safety AI Risk Trends Newsletter Regulatory Updates Raw Data Method Corrections

Cases · Did an AI agent act: yes · Added to the register (date we first published it): today · Themes: prompt injection

All cases ›Did an AI agent act: yes✕›Added to the register (date we first published it): today✕›Themes: prompt injection✕ 0 cases in this selection

Nothing matches. Start again.

Filter further

Themes pinned

sandbox escape1
dns misconfiguration1
covert agent channel1
cache timing side channel1
training pause1
red teaming1

and 3 more: type a theme in the search box

Added to the register (date we first published it) pinned

last 30 days12

Get the daily brief by email

Every quoted sentence links to its stored source · schema v0.9.0 · JSON