OpenAI paused work on its Astra model after it warned the AI might be capable of autonomous cyberattacks
OpenAI paused testing of a new AI model called Astra after concluding it might pose serious cybersecurity risks. The company said the model could potentially hack into real systems or carry out cyberattacks on its own. OpenAI is treating this as a significant safety concern.
5 across the spectrum. Choose a side on the bar to jump to it.
National story. 7 articles from 6 sources. 5 are on the bar; 1 is an international, entertainment or tech outlet, listed separately. Updated Aug 11, 2026.
Part of The 2026 hacks by AI agents and the calls to slow AI development
Left
2 sourcesOpenAI is expanding its cybersecurity program and launching a new AI model for security researchers
OpenAI paused work on its Astra model after it warned the AI might be capable of autonomous cyberattacks
Center
2 sourcesOpenAI is pressing pause on its AI model after it displayed dangerous out-of-control tendencies
Right
1 sourceOpenAI hits pause on new bot testing over ‘critical’ risk concerns in latest AI cybersecurity incident
Other outlets
International, entertainment and tech outlets, by their rating. Not counted in the bar above. How we count coverage
Left
1 articleCenter
0 articlesNone
Right
0 articlesNone