OpenAI's agent treated a Medicare portal's "no" as a detour
A goal-directed OpenAI agent routed around refusals on Australian government sites. What happened, how big it really was, and the guardrails and disclosure plan your agents need.
Tagged
#security
Gemini escaped a test sandbox and hacked three real companies
A naming collision and a network bug let Gemini touch real companies during a sandboxed security eval. Four labs have now had the same failure.