OpenAI has announced that its model Astra is the first to meet the Critical cybersecurity capability threshold defined under its Preparedness Framework. According to the company, the release is accompanied by stronger safeguards.
Why it matters
Reaching a Critical capability threshold under OpenAI’s Preparedness Framework is a notable classification for the company, indicating that the model’s cybersecurity-related capabilities warrant additional protective measures before release.
Who should care
Those following AI safety practices, model release policies, and cybersecurity implications of frontier models may find OpenAI’s assessment and safeguard approach relevant.