AI leaders’ call to slow frontier models sparks backlash

Senior figures at Anthropic, OpenAI, and other AI groups urged a slower pace for “frontier” AI—the most capable systems—arguing that safety measures have not kept up with what the technology can do. That put an industry usually associated with racing ahead into a more basic argument: whether the greater danger is deploying powerful software too quickly or falling behind competitors. Donald Trump rejected the case for new guardrails, while Chinese officials also criticized slowdown proposals; the debate spread through Washington and Beijing as the United Nations called for urgent action. The dispute pressured technology shares by adding the possibility that returns on enormous AI investments could take longer to arrive.