Astra

Astra is an AI model that OpenAI is building but hasn't released to the world yet. It's currently in internal testing, where its capabilities and risks are being measured together before any public launch.

It's a bit like a new drug showing unexpectedly strong effects in clinical trials before release, prompting extra safety checks. Astra reportedly showed a major leap over previous models in its ability to write and run code on its own, and in finding weaknesses in security systems. But because those abilities are so strong, there's concern that if misused, they could cause serious damage to real computer systems.

So OpenAI has strengthened safeguards ahead of release, even though the risk rating hasn't been finalized. In short, Astra isn't a model that's been confirmed as dangerous — it's a model being handled more cautiously because it could be.

News coverage describes it as "Astra's cyber capabilities cannot rule out a 'critical' threshold." It's easy to misread this as meaning an actual incident or leak involving Astra has occurred — it hasn't. OpenAI has stated clearly that Astra is still an unreleased model undergoing benchmarking and evaluation, and that it is unrelated to the separately reported Hugging Face security incident from around the same time.