OpenAI flags critical cybersecurity risks in Astra model, halts some development work
OpenAI has flagged potential critical cybersecurity capabilities in its forthcoming Astra AI model, leading the company to halt certain internal development activities and implement enhanced security measures.