OpenAI, Anthropic ask U.S. government to consider slowing down AI

OpenAI and Anthropic have endorsed a letter signed by hundreds of AI company employees asking the U.S. government to consider slowing down AI development if it becomes too rapid. The letter does not specify concrete mechanisms for slowing development but asks policymakers to evaluate the pace of progress and intervene if necessary.

The move marks a notable shift, as both OpenAI and Anthropic had previously focused on self-regulation and internal safety measures. The companies have faced criticism for releasing increasingly capable models without comprehensive safety testing. By endorsing the letter, they signal agreement with employee concerns about the potential risks of rapid AI advancement. The letter's signatories include engineers, researchers, and other staff from multiple AI firms.

Congress has held hearings but not passed significant legislation. Some experts argue that formal government oversight is necessary to prevent negative outcomes, while others warn that excessive regulation could stifle innovation. The letter's impact on policy remains unclear.