Eight leading AI companies, including OpenAI, Meta, Anthropic, and DeepSeek, do not have credible plans to prevent catastrophic AI risks, a new study shows. Continue reading...