Politics

Copy article

US Senator Opens AI Talks With Beijing on Capitol Hill. Britain's Multilateral Framework is Threatened

ended 01. May 2026

Senator Bernie Sanders hosted Chinese AI governance officials from Beijing's Ministry of Science and Technology on Capitol Hill to discuss AI cooperation, bypassing the long-standing multilateral framework the UK spent three years brokering at Bletchley, Seoul, and Paris.

On 29 April Senator Bernie Sanders hosted a panel with Xue Lan of Tsinghua University and Zeng Yi of the Beijing Institute of AI Safety and Governance, both tied to China's Ministry of Science and Technology, alongside MIT's Max Tegmark. The topic was AI existential risk and the need for US-China cooperation. 

Two days earlier, Treasury Secretary Scott Bessent attacked the event publicly. The same week, a Bull Moose Project report covered by the NY Post claimed a donor network centred on Dustin Moskovitz's Coefficient Giving has spent $5.9 billion on AI safety causes and pledged $37.8 billion more, with $611 million going to US political candidates. The NY Post framed the whole thing as an American fight between "doomers" and innovators, with Anthropic's $20 million political fund on one side and a $125 million deregulation fund on the other.

The pro-AI faction believes US regulation hands China the advantage in the AI race, and that safety advocates are funded obstacles to American competitiveness.

For Britain, this matters in ways the American coverage completely ignores.

The UK built an entire diplomatic position on being the convening power for international AI safety at Bletchley in 2023, co-hosting Seoul in 2024, chairing the international network of AI Security Institutes. China signed the Bletchley Declaration but refused to sign the Seoul Declaration. The US and UK both refused to sign the Paris AI Action Summit statement in February 2025. 

The multilateral process the UK anchored its strategy to is fragmenting. Now a US senator is opening bilateral AI governance talks with Beijing officials on Capitol Hill, outside of any of those pre-existing frameworks.

Meanwhile, Liz Kendall's speech on 28 April acknowledged that 70% of global AI compute is controlled by five companies, all American, and called for Britain to stop being a passive deployer. 

The UK's AI Security Institute runs on £66 million a year. American political campaign funds spent three times that in a single month on advertising about whether AI should be regulated at all. The UK still has no AI-specific legislation.

We'd like your views:

  • The UK built its AI strategy around multilateral safety cooperation, Bletchley, Seoul, the AISI network. If the US and China are now opening bilateral AI governance conversations outside those frameworks, what is Britain's convening role actually worth? (Attendees: Australia, Brazil, Canada, Chile, China, EU, France, Germany, India, Indonesia, Ireland, Israel, Italy, Japan, Kenya, Saudi Arabia, Netherlands, Nigeria, Philippines, Korea, Rwanda, Singapore, Spain, Switzerland, Turkey, Ukraine, UAE, UK, US)
  • Liz Kendall says 70% of global AI compute sits with five American companies. When the regulatory environment those companies operate under is being shaped by hundreds of millions in US political spending, what does "responsible AI adoption" mean for a British SME building on their infrastructure?
  • China signed the Bletchley Declaration but not Seoul. The US and UK refused to sign Paris. Sanders is now hosting Chinese AI officials on Capitol Hill. Is the international AI governance framework the UK invested in still functioning, or are we watching it quietly collapse and the UK's grip on AI failing?
  • The Lydia Moynihan NY Post article frames AI safety advocacy as a "doom machine." The UK government funds its own AI Security Institute to do safety testing. Are these the same thing, and should British media be importing the American political framing without examining who benefits from it? 

3 responses from the Newspage community

Copy all

Star Quote
Copy

Colette MasonVerified
AI Ethics Consultant at Clever Clogs AI
Sanders isn't just talking. He and Ocasio-Cortez introduced the Artificial Intelligence Data Center Moratorium Act to freeze US data centre construction. Meanwhile Anthropic spent $20m backing pro-regulation political candidates and the opposing camp raised $125m for deregulation. These factions have the power to control everyone's future. It is worrying that two countries have broken away to set the agenda.

The UK's AI Security Institute runs on £66m a year. Pro-AI US factions burned through more than that in a single month on lobbying and advertising.

The question for British business leaders, regulators, and policymakers isn't which side of the US argument is right. It's what happens when the country that positioned itself as the honest broker on AI safety discovers that the two biggest players are having the conversation without it.

While politicians stroke their chins, the nimble private sector is running rings around them letting multiple irreversible genies out of the bottle.
Copy

Editor
Katrina YoungVerified
Futurist | Chief Technology Officer CTO/CDO Ethical AI & Digital Transformation Strategist at KYC Digital
The Bletchley Summit gave the UK convening power, not control. China signed Bletchley, not Seoul. The US and UK declined Paris. The framework Britain built is not collapsing, but it is being bypassed in critical moments by bilateral talks and domestic political agendas. Meanwhile, 70% of global AI compute sits with a handful of US firms. British businesses are expected to practise responsible AI on infrastructure they don’t own, under standards they didn’t set. That is the asymmetry. Governance responsibility without infrastructure leverage is liability without recourse. The UK’s role still matters, but only if it converts convening into procurement power, enforceable standards, and strategic control over infrastructure.
Copy

Rohit Parmar-MistryVerified
Founder at Pattrn Data
Britain's convening role only has value if it creates use. If serious governance conversations are drifting into direct US China channels outside the Bletchley, Seoul and Paris process, that is a warning that the UK may have been hosting the discussion without shaping the power behind it. Summit diplomacy matters, but only if it changes the behaviour of the states and firms that control compute, capital and deployment at scale.

For British businesses, this makes responsible AI adoption harder. If most core infrastructure sits with a handful of US firms operating inside a heavily politicised regulatory environment, UK SMEs are inheriting governance risk they do not control. That means assurance, auditability and supplier dependency matter more than innovation branding.

The uncomfortable truth is that multilateral AI governance is not dead, but it is clearly weaker. Britain now needs influence over infrastructure, standards and procurement, not just another seat at the table.