The AI regulation smackdown isn’t over

At the start of the week, a number of high-profile AI executives appeared to converge around the idea of slowing and better governing advanced AI. The prompt came from Anthropic CEO Dario Amodei, who set out a three-step approach that would place third-party evaluators inside labs, coordinate safety measures across the domestic industry, and seek international agreements possibly with government assistance. Executives including OpenAI’s Sam Altman, DeepMind co-founder Demis Hassabis, and SpaceX’s Elon Musk publicly acknowledged agreement with elements of that plan.

Anthropic and OpenAI had previously dropped signals that they and other labs were exploring an industry framework to manage risks, a stance that could serve both reputational aims and a response to a string of concerning security incidents. OpenAI’s global affairs head, Chris Lehane, told The Verge that many people want assurances that AI is developed safely and argued that government should play a role; he said a policy window exists with growing bipartisan support for mandatory national safety standards for frontier AI.

Despite those signs of alignment inside parts of the industry, meaningful government involvement looks uncertain. The Trump administration has publicly dismissed fears about an AI safety crisis as a “hoax,” and the president sought to reassure audiences by telling Nvidia CEO Jensen Huang on a conference call that 'The robots will not be taking over.' That stance sits uneasily with earlier Trump-era actions on AI policy: an executive order from the first Trump administration, Executive Order 13960, established principles for federal AI use and remains on the books.

The AI world itself remains divided. Meta CEO Mark Zuckerberg quickly opposed limits on company autonomy and, according to a Wall Street Journal report, he along with Elon Musk and Nvidia’s Jensen Huang blocked proposals for an industry-funded independent regulator modeled on financial-sector groups like FINRA. Some companies declined to comment to The Verge, while others reiterated positions that labs bear responsibility for training models safely. Analysts and former government officials said the administration’s shifting commentary has produced an inconsistent public posture on regulation and that debates over the nature and scope of guardrails are far from settled.