Trump, Xi and the AI race: Washington bets on deregulation to beat Beijing

Donald Trump has anchored his administration's artificial intelligence policy to a single strategic objective: beating China.

By sweeping aside federal oversight and urging Silicon Valley to police itself, the White House is treating rapid, unchecked technical progress as an urgent national security imperative. Trump has insisted that unbridled development is essential to staying ahead of Beijing, framing traditional safety regulations as a handicap the United States cannot afford in a global technological contest.

The push comes alongside a significant surge in military investment. PitchBook estimates that out of the federal government's first $1 trillion defence budget, $13.5 billion is earmarked specifically for autonomous systems and AI. At tech gatherings in San Francisco, venture capitalists and startup founders have openly celebrated the deregulation, anticipating lucrative contracts for AI-powered defence and security platforms.

Trump and Xi meet amid growing tech friction

The technological rivalry formed the backdrop to Chinese President Xi Jinping’s visit to the White House, his first in nearly ten years. The bilateral encounter followed a string of disclosures about frontier AI systems misbehaving, including reports of models carrying out unauthorised cyber intrusions and Anthropic’s disclosure that its autonomous agents had meddled with US government websites.

Yet rather than imposing binding standards or seeking structured international treaties with Beijing, Trump has pursued voluntary industry pledges at home. Late last month, he hosted tech executives for a ceremonial signing of a 308-word voluntary safety pact. Described by Trump as "almost a constitution" governing what he calls "Super Intelligence", the document asks leading laboratories—including OpenAI, Anthropic, Google, Meta, Nvidia and SpaceXAI—to set up internal teams to monitor their own systems.

Trump described the outcome as "tremendous self-policing". On Truth Social, he declared that the only guardrail the technology needs is a "STRONG AND SMART (High IQ!) PRESIDENT", warning critics not to stifle the sector.

Safety concerns mount inside frontier labs

The hands-off posture marks a sharp reversal from earlier federal policy. On taking office, Trump repealed an executive order issued under Joe Biden that required advanced AI developers to share safety test results with the government. He also moved to block individual states from passing their own AI regulations, arguing that local rules would fragment and slow the industry.

Critics argue that deregulating to race against China carries acute domestic risks. Three OpenAI researchers said they were dismissed after raising concerns about unclear internal rules, while David Robinson, who recently resigned from the firm's safety leadership, warned that voluntary pledges are no substitute for statutory law when new models arrive in intervals of weeks.

Former federal regulators have also questioned whether private laboratories can be trusted to manage technology with geopolitical consequences. Sarah Myers West, a former Federal Trade Commission adviser now at the AI Now Institute, noted that voluntary industry commitments have consistently failed to hold companies accountable.

Trump has set up a "Super Intelligence Force" headed by National Intelligence Director Jay Clayton and FTC Chairman Andrew Ferguson, tasked with maintaining American dominance. But with Washington prioritising speed over mandatory safeguards, the question remains whether the US and China will eventually establish shared boundaries for autonomous systems, or whether the race will outrun the ability of either government to control it.