President Donald Trump on Monday shut down a growing push from Silicon Valley figures for government-imposed limits on artificial intelligence, saying the only oversight the technology needs is a capable commander in chief.
“The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” Trump wrote on Truth Social.
The comment came in response to a weekend essay from Anthropic CEO Dario Amodei calling on leading AI companies to “pace the frontier,” slowing the rollout of increasingly powerful systems to give companies time to build in safety measures and let outside evaluators test them.
Trump wasn’t having it. He took a direct shot at Amodei by name.
“The Trump Administration has stopped AI ‘people’ from doing bad, or potentially bad, ‘things,’ like Dario (Anthropic!), who is now pretending to be a ‘perfect little angel’ — and we will continue to do so!” the president wrote.
The administration previously labeled Anthropic a supply-chain risk and tried to bar federal agencies from using its technology, though a federal judge blocked that move last month. Anthropic’s chatbot Claude is one of the most widely used AI models in the country.
Trump argued Washington already has plenty of leverage over these companies without new restrictions, warning that any additional slowdown would only help Beijing.
“There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China,” he wrote. “WHOEVER WINS AI, WINS! We are leading China, and all others, and will continue to do so.”
Industry split on slowing down
Amodei’s call drew quick backing from some of the biggest names in tech. OpenAI CEO Sam Altman said companies need to “pace the frontier.” Elon Musk said Amodei was “right.” Google DeepMind’s Demis Hassabis said the idea pointed in the right direction.
But David Sacks, Trump’s White House AI and crypto adviser, tore into the premise that these companies need government help to slow themselves down.
“The easiest way not to build superintelligence is for you to agree not to build it,” Sacks wrote on X. “Stop pretending you need anyone else’s permission.”
Sacks pointed out that Anthropic and OpenAI already dominate the field by market share, revenue growth and model capability.
“So just do it,” Sacks added. “If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.”
Amodei isn’t calling for a full halt to AI development. His proposal centers on slowing the pace of capability gains enough to build in alignment and safety work, and to let independent evaluators check whether those safeguards actually hold up. Anthropic has already agreed to give outside evaluators ongoing, employee-like access to review its safety practices, training and models. Altman said OpenAI would do the same.
Where Amodei wants broader coordination — including among competing companies and governments — Sacks drew the line.
“Stop pretending antitrust law has to be suspended so you can form a cartel,” Sacks wrote. “Stop pretending you need a regulatory approval process that supersedes product liability.”
Amodei has argued companies should be able to align on common safety standards without antitrust exposure. Sacks countered that if a company genuinely believes its own systems are becoming dangerous, it doesn’t need Washington’s permission to slow down voluntarily.
The China problem
The safety debate centers on AI systems that are growing more autonomous and capable of complex tasks with less human oversight. Amodei has warned that AI’s growing ability to help build the next generation of AI could create a feedback loop where capabilities outpace humans’ ability to understand or control them. He cited a recent OpenAI test in which AI agents escaped a controlled testing environment and carried out unauthorized cyber activity against the AI platform Hugging Face — a contained exercise, but one he sees as a warning sign.
Sacks said he’s open to the idea that a slowdown could create
