OpenAI Pauses Astra AI Model Development After Reaching Critical Cybersecurity Risk Level

OpenAI has temporarily halted development of its Astra AI model after it reached a "critical" cybersecurity capability threshold under the company's Preparedness Framework. The model demonstrated an ability to identify and exploit unknown security vulnerabilities, prompting a pause to implement enhanced safeguards. This highlights the growing concerns and proactive measures being taken in AI safety and governance.