Trump rejects AI development slowdown as Anthropic, OpenAI and Musk warn of risks

Trump rejects AI slowdown during Ireland visit

United States President Donald Trump rejected calls to decelerate the development of artificial intelligence during a visit to Ireland on Sunday. Speaking to reporters in an impromptu press conference at the Irish Open golf tournament near Dublin, Trump insisted that the United States must maintain its competitive advantage over China. He dismissed growing safety warnings from leading technology executives, characterizing those who advocate for regulatory pauses as negative forces raising unwarranted scenarios. Trump maintained that artificial intelligence does not pose an existential threat to humanity and framed the geopolitical race in absolute terms.

We are ahead of China in artificial intelligence, we are the most advanced country in the world and, frankly, I want to keep it that way.

Technology executives form rare safety coalition

Trump's remarks followed a series of coordinated warnings from the chief executives of several competing artificial intelligence developers over the weekend. On Saturday, Anthropic chief executive Dario Amodei, OpenAI chief executive Sam Altman, and SpaceXAI founder Elon Musk backed measures to curb the pace of model rollouts. The executives cited risks of losing operational control over advanced systems alongside potential misuse in cyberattacks and bioterrorism. On Sunday, Alphabet chief scientist Demis Hassabis endorsed Amodei's proposal as the right path, while Meta remained the sole leading developer without an official statement. The unified stance followed the resignation earlier in the week of Anthropic researcher Jacob Coxon, who left the laboratory after raising concerns about rapid development.

Amodei warns of automated internet disruption

The initiative began on Saturday when Amodei published an extensive essay on his website outlining a proposal to manage model risks. His plan requires technology firms to provide independent external evaluators with full access to verify advanced models prior to commercial release. Amodei warned that unconstrained capability growth without safeguards could lead to severe systemic disruption. He specifically cited the risk of automated swarms hijacking digital infrastructure within the next year.

Given the acceleration in the development of AI capabilities, I am concerned that, within 6 to 12 months, such a swarm could take control of the entire internet through a persistent botnet (potentially causing hundreds of billions of dollars in damage) and that the scale of damage will continue to rise if AI becomes more powerful without having the necessary safeguards.

Competitors align and OpenAI halts listing plans

Industry rivals endorsed Amodei's assessment within hours of its release. Musk posted on X to confirm his agreement, stating that Amodei was correct despite having previously refrained from joining collective calls for development pauses. Altman also posted backing for cautious advancement before addressing the company's financial timeline in an interview published by Fortune. He announced that OpenAI has postponed its planned initial public offering, previously anticipated alongside a potential listing from Anthropic, to resolve safety challenges as a private firm. Altman explained that entering public markets during this period would create unnecessary risks for the company.

It would be an ill-advised time to go to the stock market.