Model Release Gating

OpenAI says its next model finds and exploits unknown flaws without a human in the loop

Astra is the first model to trip the cyber threshold written into OpenAI's own safety protocol, so security teams inherit the offensive uplift and a set of guardrails the company admits may stop legitimate work.