The Bosses of AI Want to Slow Down. Trump Says No
The people building AI are now asking the world to hit the brakes.
Anthropic CEO Dario Amodei published a long essay called "We Must Pace the Frontier," arguing that AI should be developed more carefully, not faster. He proposed three steps: outside experts allowed inside AI companies to check safety claims and report problems, cooperation between big AI labs in democratic countries on shared rules, and eventually talks between democracies and authoritarian governments. Anthropic says it will start with the first step on its own. Amodei stressed that "pacing" does not mean stopping — just going slower than they otherwise would.
What made this unusual is who agreed. OpenAI's Sam Altman said he agrees, and that OpenAI would also welcome independent evaluators and a federal safety framework. Google DeepMind's Demis Hassabis called the direction correct. Elon Musk simply posted: "Dario is right." These are direct competitors who rarely line up publicly, which suggests the industry is nervous about how fast things are moving and about who will get blamed if something goes wrong.
President Trump pushed back hard on Truth Social, writing that the only control AI needs is a "strong and smart (High IQ!) President," that his administration already has "tremendous criminal and regulatory power" over tech companies, and that "WHOEVER WINS AI, WINS." He called criticism of AI a "sick conspiracy" that only helps China. Nvidia CEO Jensen Huang, whose chips power most AI, took a friendlier line, saying everyone will win and "the robots will not be taking over." Vice President JD Vance said it was "a little bit weird" for AI companies to come to government asking for rules.
So what actually changed? No law passed. No AI model was slowed down. These are voluntary promises from companies, and the US government is signaling it will not force anyone to keep them. But this fight decides two things that touch your life: whether anyone outside a company is allowed to check what AI can do, and whether America or China writes the rules everyone else follows.
- The heads of Anthropic, OpenAI, Google DeepMind and Elon Musk all publicly backed slowing AI down and letting outside experts audit their work.
- President Trump rejected the idea, saying AI needs no guardrails beyond a strong, smart president — and that beating China matters more.
- No AI model has actually been paused; these are voluntary company promises, not laws with penalties behind them.
Why It Matters
Who gets to put the brakes on AI — and whether anyone outside checks it — shapes your privacy, job, and safety.