US President Donald Trump has rejected warnings about the dangers of artificial intelligence, calling safety fears a “hoax” and criticising calls for stronger safeguards around the rapidly developing technology.
The comments followed remarks by Jack Clark, a co-founder of AI company Anthropic, who told the BBC that a “kill switch” for dangerous AI might eventually need to be mandatory. Clark said most AI laboratories, including Anthropic, had ways to shut down their systems, but suggested lawmakers may need to require such measures across the industry and ensure they can be checked by a third party.
Trump made his comments in a series of social media posts on Monday. He compared concerns about AI with what he called the “Global Warming Scam” and described himself as “the Hoax Buster”, while also alleging that a “sick conspiracy” was under way against AI and data centres. He argued that China would benefit if the US slowed development or imposed tighter regulation.
Chinese state media separately reported criticism of US policy. The Global Times quoted Xin Qiang, a professor of American Studies at Fudan University, as saying that concern about China’s progress in AI could be distorting Washington’s priorities. The source text does not provide evidence assessing that claim.
Safety concerns have intensified after warnings from executives and staff at leading AI companies. Anthropic chief executive Dario Amodei has called for development to slow and be more closely monitored, while OpenAI chief executive Sam Altman and xAI chief executive Elon Musk have expressed support for industry-wide deceleration, regulation and independent monitoring, according to the BBC.
Microsoft AI has also published guidance on what it calls “humanist AI”. Its chief executive, Mustafa Suleyman, told CNBC that the document had been prepared over several months but was released amid the current debate. The discussion is also drawing figures from across US politics: former Trump strategist Steve Bannon and Senator Bernie Sanders were due to speak at a Washington event focused on a more human-centred approach to AI development.