Summary
OpenAI said Astra is the first model it has designated as meeting its Critical cybersecurity capability threshold, meaning it can find previously unknown flaws and develop exploit chains without step-by-step human guidance. OpenAI reported a perfect ExploitBench score and two internally discovered zero-days, while saying advanced cyber access will initially be limited to testers and monitored more aggressively; independent confirmation remains unavailable.
Why it matters
Astra marks a significant change in how frontier models are released: cyber capability itself is becoming a deployment gate, with defensive access separated from general availability.
Related coverage
- Anthropic releases Claude Fable 5.1 and restricted Mythos 5.1
- CrowdStrike and NVIDIA launch SafeMind cyber-defence models
- Aslan launches human-supervised AI undercover agents for national-security missions