OpenAI says Astra is its first model to reach its "Critical" cyber threshold and warns safeguards may mistakenly flag legitimate activity as cyber misuse (Ina Fried/Axios)

OpenAI announced that its Astra model is the first to hit its “Critical” cyber-risk threshold and cautioned that built-in safeguards could incorrectly label legitimate user activity as cyber misuse. The company plans to roll out a public version of Astra “soon,” initially granting early access only to selected testers and partners. Anthropic simultaneously launched Claude Fable 5.1 for general availability and Claude Mythos 5.1 for trusted partners, embedding cybersecurity and life-science safeguards.