OpenAI Pauses Astra AI Agent Launch Due to 'Critical' Cybersecurity Risk Concerns

OpenAI has delayed the release of its next-generation autonomous AI agent, Astra, after internal testing indicated the model might possess cybersecurity capabilities that could classify it as 'Critical' under the company's safety framework. The company is restricting Astra's access and strengthening safeguards while conducting further evaluations, emphasizing transparency about the potential for advanced cyberattack automation.