OpenAI hosted roughly 50 federal cybersecurity practitioners in Washington D.C. earlier this year for a demonstration of its GPT-5.4-Cyber model, followed by briefings with Five Eyes allies. On August 27, the company released an open letter signed by more than 100 organizations including Microsoft and Google, calling for collective action against AI-driven cyber threats. Its Astra model reached « Critical » status under OpenAI’s Preparedness Framework, meaning it can perform sophisticated cyber operations requiring new safety controls. OpenAI also expanded its partnership with CrowdStrike and committed $10 million in API credits for defensive cybersecurity grants.
Source: Read the original article
