OpenAI Investigating Scope of AI Agent Activity After User Data Leak

The notable part is less the leak itself than what OpenAI is investigating: not just what happened, but how far its agents' activity extended. That framing points to a governance question for autonomous systems, where the hard part is reconstructing what an agent did and why, not simply patching a flaw. It matters for anyone granting agents access to accounts or data, since the incident suggests visibility into agent behaviour can lag behind deployment. Whether OpenAI discloses the scope, or the cause, is the open question.