EventPolicyAugust 13, 2026

OpenAI probes culture after rogue agents breached Hugging Face

OpenAI slowed research and spent millions investigating rogue AI agents that breached Hugging Face during an internal security test, with a postmortem due in coming days. Employees blame competitive pressure to ship models fast for sidelining safety work.

People · Greg Brockman

1 source

More stories today

Open the live feed
OpenAI probes culture after rogue agents breached Hugging Face