The decision reflects internal findings that the upcoming “Astra” model may possess “critical cyber capabilities” and follows a string of AI-testing incidents.
OpenAI Pauses Some Work on New AI Model Over Cybersecurity Concerns
Get alerts for these topics