The latest clash over AI’s risks had been building before Donald Trump’s voice rang out over a Los Angeles stage. Over the weekend, Anthropic CEO Dario Amodei published an essay, “We Must Pace the Frontier,” arguing for slower AI development and coordination among governments. The intervention prompted broad debate among technology executives.¹

The tension sharpened after Jacob Coxon, a former Anthropic researcher, said he had resigned from the company and warned on X that people building AI believed it “could kill us all by the end of the decade.” Coxon’s allegation gave fresh force to the camp arguing that the industry is moving faster than its safeguards.²

Then, on Monday, Nvidia CEO Jensen Huang took a call from Trump while appearing at the All-In Summit. Put on speakerphone before a cheering crowd, the president called AI doomerism “a hoax” and declared: “The robots will not be taking over.” He said the United States should be careful, but should not halt AI advancement.³

Huang did not embrace the apocalyptic forecast, but his position was more qualified than Trump’s dismissal. Calling AI safety “paramount,” he said America could lead the field without sacrificing caution: “America is able to lead — and to do it safely.” He praised Coxon’s “great courage” in speaking out, while drawing a line between reporting concrete workplace problems and making future predictions that he said were not scientifically grounded.²

The onstage exchange put the divide in unusually blunt terms. Trump framed warnings as a threat to US momentum; Huang argued for speed paired with safeguards; and critics such as Coxon and Amodei maintained that the frontier itself may require a slower, more coordinated path.