Anthropic model takedown fuels warning of ‘ad hoc’ AI regulation

OpenAI pauses development of Astra AI model over critical cybersecurity capability concerns