The world's most prominent artificial intelligence labs are calling for new government regulations to prevent machines from evolving beyond human management. A coalition of researchers, including leadership from OpenAI, Anthropic, Meta, and Microsoft, has urged officials to probe autonomous research to avoid a loss of control that could lead to the marginalization or extinction of the human race.

The group argues in a paper, first reported by the Wall Street Journal, that self-improving systems may soon compress years of technological development into mere months. This acceleration would make it impossible for regulators to keep pace with AI that can iteratively refine its own capabilities without human intervention.