Anthropic CEO Dario Amodei published a lengthy essay on Saturday arguing that AI development should slow down, and the response from Silicon Valley and Washington has been swift and varied. Sam Altman, Elon Musk, and several politicians weighed in, while US President Donald Trump took the opportunity to rant against the whole idea.
Amodei’s essay, titled “We Must Pace the Frontier,” made three main proposals: embedded third-party evaluators to check company safety practices, coordination between frontier AI firms in democratic countries, and global coordination between those governments and authoritarian ones. He said Anthropic is “unilaterally committing to the first of these steps.”
What Amodei Actually Said
Amodei was careful to define his terms. He wrote that pacing “does not mean halting model training or technical progress.” Instead, he said companies should take adequate time to align and safeguard their models, with third-party evaluators confirming that work.
His essay came as part of a broader conversation about the risks of AI. Amodei framed the question as one of balance: how to move forward without letting the technology outpace the ability to understand it.
Altman Agrees, Sort Of
OpenAI CEO Sam Altman responded quickly, posting on X that he agreed with Amodei. “This has been a primary topic of discussions we’ve had at OpenAI in recent weeks,” Altman said.
He also committed to independent evaluators at his own company, calling it a “great idea.” On Sunday, Altman went further, welcoming “a federal framework that sets consistent safety requirements for frontier AI.” He urged other companies to learn from OpenAI’s approaches and proposed shared standards for misalignment, monitoring, and safety.
Like Amodei, Altman stressed that pacing does not mean stopping. Progress should simply be “slower than it otherwise could be.”
Musk Nods, Trump Nails It
Google DeepMind co-founder and chair Demis Hassabis replied warmly to Amodei’s essay. “The details need working through, but the direction is correct for meeting this critical moment,” Hassabis said.
Elon Musk agreed more directly. “Dario is right,” Musk posted, sharing Amodei’s essay.
Trump took a different approach. His Monday morning Truth Social post was a rambling mix of policy argument, personal attack, and capital-letter shouting. He dismissed the idea of guardrails, insisting that the “STRONG AND SMART (High IQ!) PRESIDENT” is the only control AI needs. He mocked Amodei as someone “now pretending to be a perfect little angel” and warned of a “SICK conspiracy going on against AI and Data Centers.”
Vance’s Skepticism
US Vice President JD Vance pushed back against the idea that companies are coming to the government to beg for regulation. He called it “a little bit weird” and said it “feels a little bit to me like a bit of a trojan horse.”
He also pointed to the President’s framing of the competition with China. “We know the Chinese are engaged in this long term AI race against the United States,” Vance said.
Johnson Wants a Meeting
House Speaker Mike Johnson said in a Sunday interview with Jake Tapper on CNN that he wants frontier AI companies to meet with Congress. “They need to come together in a meeting with us and sit down and figure out what the right balance is,” Johnson said.
He ruled out a moratorium. “We cannot put a moratorium on this because China will overlap us,” he said. But he also emphasized national security and model safety. “It’s national security balanced with the immediate security of making sure the models are safe,” he said.
It is unclear if or when that meeting will happen. Johnson said he would do it “tomorrow.”
Sanders Agrees With Noonan
Senator Bernie Sanders aligned himself with a surprising ally: Peggy Noonan, the Wall Street Journal columnist and former Ronald Reagan speechwriter. Sanders said he agrees with Noonan’s article titled “Pause AI for Humanity’s Sake.”
| Speaker | Position |
|---|---|
| Dario Amodei | Published essay calling for pacing |
| Sam Altman | Agrees with Amodei, supports independent evaluators |
| Donald Trump | Rejects pacing, mocks Amodei |
| JD Vance | Calls company requests for regulation “a bit of a trojan horse” |
| Mike Johnson | Wants a meeting, rejects moratorium |
| Bernie Sanders | Agrees with Noonan’s call to pause |
Reading Between the Lines
The split between Trump and the rest of the field is stark. Every other speaker treated the issue as a serious technical problem that requires coordination. Trump treated it as a political cudgel.
His post is the most entertaining part of an otherwise dry debate. It is also rambling and dismissive. Trump insists the President is the only control AI needs while warning of a conspiracy against data centers and China. He threatens “Conspiracy Theorists, Treasonists, Traitors, and Leakers” without explaining what any of those categories mean in the context of AI policy.
The comparison between Trump and the rest of the field shows how deeply the political system is still grappling with the technology. CEOs and politicians alike are talking about coordination and standards, but the President is offering a strong leader as the only solution.
What Comes Next
Amodei’s essay set off a chain reaction. Altman committed to evaluators, Musk agreed, and Trump raged. Vance and Sanders raised questions about the motives behind the push for regulation. Johnson wants a meeting.
Whether any of this leads to actual action is unknown. Amodei’s unilateral commitment is a start, but the coordination he wants requires everyone else to follow.
The debate is likely to continue. The conversation around AI safety remains unresolved.
Get the Notebook.
The day's best stories and every fresh verdict, in plain English, in your inbox by seven. One email a day, no more.