OpenAI's Astra is first model to hit Critical cyber threshold; public release planned
OpenAI announced that Astra is its first model to meet the Critical cybersecurity capability threshold under its Preparedness Framework, with stronger safeguards for release. The company plans to publicly release a version of Astra soon, but will limit advanced cyber capabilities to select partners, and warns safeguards may mistakenly flag legitimate activity as misuse. Separately, The Information reports Astra uses a technique called 'recurrent depth' that improves cost and performance but obscures reasoning, making monitoring harder.
Coverage timeline
OpenAI Blog
Astra is the first OpenAI model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, with stronger safeguards for release.
Techmeme
Wired : OpenAI says it plans to publicly release a version of Astra “soon” but will make its advanced cyber capabilities available only to select partners — The company will give select partners early access to its Astra AI model—so they have time to shore up their defenses.
Ina Fried / Axios : OpenAI says Astra is its first model to reach its “Critical” cyber threshold and warns safeguards may mistakenly flag legitimate activity as cyber misuse — OpenAI said Tuesday that it plans to release its latest model — Astra — soon, but its most advanced cybersecurity features …
The Information : Source: OpenAI's Astra model uses “recurrent depth”, a technique that improves cost and performance but obscures the AI's reasoning, making it harder to monitor — OpenAI says its forthcoming AI model Astra marks a step up in capabilities such as coding and operating applications on a computer.
