AI labs intensify internal debate over superintelligence doomsday risks

Researchers at Anthropic, OpenAI, Meta, and Google are escalating internal discussions about existential risks from superintelligent AI, signaling a shift toward treating these concerns as credible and urgent. The article reports on private deliberations, but specifics on proposed safeguards or timelines remain unclear. This matters because it could presage changes in AI development practices, regulation, and public messaging from leading labs.