Study finds major AI companies’ safety practices fall short of emerging global standards

A new edition of the Future of Life Institute’s AI Safety Index concludes that leading AI developers including Anthropic, OpenAI, xAI and Meta lack robust strategies to control potential superintelligent systems, leaving their safety practices "far short" of emerging international norms. The report, based on an independent expert panel, comes amid growing concern over AI‑linked self‑harm cases and AI‑driven hacking, and has prompted renewed calls from researchers such as Max Tegmark, Geoffrey Hinton and Yoshua Bengio for binding safety standards and even temporary bans on developing superintelligence until better safeguards exist.