In a striking reversal of the typical geopolitical narrative, a leading British artificial intelligence expert has told Parliament that China is currently playing the role of the “good guy” in the global development of AI, while the United States pursues a dangerous, unregulated race for supremacy.
Dame Wendy Hall, a former UN AI adviser and co-author of a foundational UK government AI review, made the assessment during an evidence session with the House of Commons business and trade committee. She stated that China is actively backing multinational efforts to establish global governance for the technology, a stark contrast to the American approach. “The US is totally against any regulation and talk about global governance,” said Dame Wendy, who is director of the Web Science Institute at the University of Southampton. “It’s all we’re going to win at all costs.”
China’s Coordinated Push
Dame Wendy noted that Chinese researchers are producing “amazing work,” are highly efficient, and show a willingness to release AI models on an open-source basis. This stance is formalised in China’s “Global AI Governance Action Plan,” which emphasises international cooperation, multilateral standards, and ethical safeguards, positioning itself as a voice for the Global South. The capability of Chinese AI is underscored by firms like DeepSeek, whose DeepSeek-V3.2 model is reported to match leading US rivals. Its upcoming model, V4, is expected to run on Huawei’s Ascend chips, signalling a deliberate shift towards China’s domestic semiconductor infrastructure and away from reliance on American components from NVIDIA and AMD.
However, this collaborative facade is complicated by Beijing’s requirement for domestic AI companies to cooperate with state intelligence work. This duality aligns with warnings from the UK government-funded Centre for Emerging Technology and Security (CETaS), which has highlighted national security risks from increasing evidence of AI collaboration between adversarial states, including China, Russia, Iran, and North Korea.
The US “Wild West”
The American stance, as characterised by Dame Wendy, is one of a deregulated “wild west,” driven by profit and hype among competing tech giants. This aligns with the “Winning the AI Race: America’s AI Action Plan” from July 2025, which focuses on technological supremacy and deregulation. More concretely, the Trump administration’s National Policy Framework for Artificial Intelligence, issued in March 2026, advocates for a “light-touch” regulatory approach, argues that AI model development is an inherently interstate activity that states should not regulate, and suggests leveraging existing agencies and industry-led standards rather than creating new oversight bodies.
This philosophy fuels the rhetoric of a direct race, as seen when former President Donald Trump claimed in January that the US was “leading China by a tremendous amount.” Yet, assessments like that from Demis Hassabis, chief executive of Google DeepMind, who said China was only about six months behind but had not yet pushed the scientific frontier, suggest a more nuanced competition.
UK Vulnerabilities and Sovereign Risks
The committee heard stark warnings that Britain’s strategic approach may be compounding these geopolitical tensions with significant domestic risk. Neil Lawrence, Cambridge University’s DeepMind professor of machine learning, cautioned that the UK’s deep reliance on US tech corporations—Google, Microsoft, OpenAI, and Amazon—risks a repeat of the Post Office Horizon scandal, where a centrally deployed system caused profound injustice.
“We’re constantly hearing about AI that works for the UK is AI that works for Microsoft, Amazon, OpenAI, Google,” Professor Lawrence said. He warned of “a lack of confidence in our own people, in our own businesses and our own universities,” arguing that “these corporations are clearly not aligned with the interests of our citizens.” When Labour MP Dan Aldridge asked if the UK had “effectively outsourced our AI model development to private billionaires, with zero loyalty to the British state and consumer,” Dame Wendy Hall replied simply: “Yes.”
This reliance is being formalised through major deals. The UK and US signed a “Tech Prosperity Deal” in September 2025, involving over £31 billion in pledged investment from US firms, including a £22 billion Microsoft commitment to build the UK’s largest supercomputer. CETaS has raised concerns that such agreements may entrench dependence on US technologies and crowd out UK companies.
Promises from these overseas giants, however, are showing signs of strain. OpenAI has placed its planned “Stargate UK” datacentre project on hold, citing high energy costs and regulatory uncertainty. A separate government plan to open “the largest UK sovereign AI datacentre” by the end of this year is far behind schedule, with the intended site still operating as a scaffolding yard.
A more fundamental obstacle is emerging: a crippling lack of power. Microsoft told the committee that a planned datacentre in the north of England would not come online until at least 2033 due to grid limitations. Kao Data, a datacentre operator, stated bluntly: “We are waiting up to 15 years now for firm grid offers.” Meanwhile, the complex risks within AI supply chains used for national security, including opaque practices and insufficient ethical due diligence, remain a pressing concern highlighted by CETaS.
