President Trump fired off two Truth Social posts Monday targeting Anthropic CEO Dario Amodei by name, calling AI safety warnings a "HOAX" and accusing the industry's own leaders of running a "SICK conspiracy" that hands China an advantage in the global AI race.
The broadside came hours after Amodei published a lengthy essay over the weekend urging the world's biggest tech companies to slow their development of artificial intelligence models. Amodei proposed a three-step plan to add guardrails to AI, a plan that drew public backing from OpenAI boss Sam Altman and SpaceX founder Elon Musk, two figures rarely aligned with the Anthropic chief on anything.
Trump was not persuaded. In his first post Monday morning, the president wrote that his administration had already reined in AI companies and would keep doing so:
"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!"
He then escalated, framing the push for AI regulation as something far more dangerous than a policy disagreement:
"We already have tremendous CRIMINAL and REGULATORY power over these companies! There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China. WHOEVER WINS AI, WINS!"
Trump closed the post with a warning: "Conspiracy Theorists, Treasonists, Traitors, and Leakers, BEWARE!"
Roughly three hours after the first salvo, Trump returned to Truth Social with a second post. This time he questioned why any industry leader would voluntarily invite regulation severe enough to destroy the business.
"Concerning AI, when, in the History of Business, did anyone see the Leaders of an Industry call for Regulation that, if strongly implemented, will drive them into oblivion and bankruptcy?"
He followed that with a flat declaration: "AI, and Data Centers, will be the Greatest Economic Development Engine in History." The president dismissed warnings that advanced AI could pose an existential threat, calling the idea a "HOAX", the same word he has used over the years to wave off claims he considers overblown or politically motivated.
The timing is not accidental. AI capital spending has become one of the few bright spots in the broader economy. The Federal Reserve has repeatedly cited massive AI-related expenditures as a highlight at its meetings, and the central bank is largely expected to hike interest rates this week. For the White House, anything that threatens the flow of AI investment threatens a talking point it badly needs.
Amodei's weekend essay did not arrive in a vacuum. The week before, AI researcher Jacob Coxon, who previously worked at both Anthropic and OpenAI, quit his job and posted on X that AI "could kill us all by the end of the decade." That post racked up more than 171 million views, turning a niche safety debate into a mass-audience spectacle.
Amodei's essay built on that anxiety. He urged the world's largest AI firms to slow their pace and outlined a three-step plan for guardrails, though the specific steps were not detailed in public reporting. He also predicted that foreign rivals may not be willing to coordinate on pacing, an acknowledgment that any slowdown by American companies could simply clear the runway for competitors abroad.
In a CBS News "Sunday Morning" interview, Amodei conceded the difficulty of getting global cooperation on AI restraint. He pointed to the military incentives that make voluntary slowdowns unlikely:
"I think that's going to be very difficult because the incentives to pull ahead and the military advantage that you get from that are so large. And honestly, I don't know if it's possible, but we should try."
That last line, "but we should try", captures the gap between Amodei's position and the White House's. Trump sees the AI race as zero-sum. Amodei sees it as a shared risk. Both acknowledge the stakes. They disagree completely on what to do about them.
The friction between the Trump administration and Anthropic predates this week. Anthropic has a history of battling with the White House over AI safety, and the company previously refused to allow the Pentagon unfettered access to its AI tools for surveillance or military capabilities. That refusal carried consequences: the Department of War labeled Anthropic a supply-chain risk, a designation typically reserved for foreign companies, not domestic tech firms.
Trump signed an executive order in June encouraging a hands-off approach to AI regulation, signaling that the administration views safety mandates as obstacles to American dominance. Anthropic's resistance to military access and its CEO's public calls for restraint put the company squarely at odds with that posture.
And the White House is not the only party unimpressed by Amodei's plea. China's response was swift and dismissive. At a summit in New Delhi over the weekend, Chinese President Xi Jinping said China would take the lead in fostering global AI collaboration among developing countries, a statement that reads less like cooperation and more like a bid for influence.
On Monday, China's Foreign Ministry spokesperson Guo Jiakun responded directly. In a statement translated by Reuters, Guo said:
"Fear mongering, confrontation, competition will just disrupt [the] process of global AI governance."
The state-backed Global Times went further, publishing an editorial that compared calls for China to slow its AI advancements to a "Cold War" tactic. Beijing's message was clear: it has no intention of pumping the brakes just because Silicon Valley executives are worried about what they built.
That reaction is the strongest evidence in Trump's favor. Whatever Amodei's intentions, the practical effect of a unilateral American slowdown would be to hand China an open lane. Beijing said so itself, in almost as many words, by rejecting the premise entirely.
Trump also took a jab at Google, which plans to build a massive AI plant in Finland because of permitting issues in the United States. The president said he wants Google "to change their thinking", a remark that underscores his frustration with domestic regulatory barriers pushing AI investment overseas.
All of this unfolds with a scheduled Sept. 24 meeting between Trump and Xi Jinping on the horizon. AI dominance is expected to be a central tension in that conversation. Trump's public posture, that any restraint on American AI is a gift to Beijing, sets up a negotiating frame in which safety advocates like Amodei become, at minimum, unhelpful to the administration's leverage.
Amodei's own concession on CBS, that military incentives make global coordination "very difficult" and that he does not know "if it's possible", undermines the practical case for the slowdown he is proposing. If even the man calling for restraint admits the other side has no reason to comply, the policy amounts to a voluntary handicap.
That does not make the safety concerns imaginary. Coxon's viral warning drew 171 million views for a reason. But warnings without workable solutions are not policy, they are atmosphere. And atmosphere does not win an AI race against a rival that has already told you it will not slow down.
When your competitor announces it will sprint and your own industry leaders ask you to walk, the president's frustration is not hard to understand. The harder question, whether the technology these companies are building actually poses the risks they claim, is one nobody in Washington or Silicon Valley has answered yet.