Key Takeaways
- The UK’s post‑World‑War‑II experience of being cut off from US nuclear cooperation is being mirrored today by potential restrictions on access to advanced AI technologies.
- A recent report by the UK Parliament’s Science, Innovation and Technology Committee warns that the country “may not be able to count on its allies” for critical AI capabilities and urges an AI sovereignty strategy.
- The Trump administration’s export ban on Anthropic’s frontier models (Fable and Mythos) demonstrated how a unilateral government decision can instantly revoke non‑US access to commercially relied‑upon AI tools.
- China is simultaneously tightening controls on its own AI firms to prevent US companies from acquiring domestic talent and IP, echoing the same protective logic that drove the 1946 McMahon Act.
- Experts liken powerful AI models to “digital nuclear weapons,” warning that unchecked progress toward artificial general intelligence (AGI) could pose existential risks comparable to pandemics or nuclear war.
- Proposals for an international treaty to halt premature development of superintelligent AI have been drafted but have yet to gain any signatories, underscoring the difficulty of achieving global cooperation in this new arms race.
Historical Parallel: The 1946 McMahon Act
Less than a year after the United States dropped atomic bombs on Hiroshima and Nagasaki, President Harry S. Truman signed the McMahon Act, ending nuclear cooperation with America’s allies, including the United Kingdom. Despite British scientists’ pivotal role in the Manhattan Project, the UK was left to develop its own nuclear deterrent at a cost of roughly £150 million—equivalent to the entire planned annual budget of the newly formed National Health Service. As the article notes, “The move triggered a nuclear arms race, costing the UK £150m – equivalent to the entire planned annual budget for the newly created NHS – at a time when the country was still struggling with rationing and post‑war reconstruction.” The success of Operation Hurricane, Britain’s first atomic test, ensured that the nation would not remain permanently dependent on Washington for its security.
Today’s Warning from UK MPs
Eighty years later, a similar scenario looms with artificial intelligence. On Tuesday, the Science, Innovation and Technology Committee released a report cautioning that the UK “may not be able to count on its allies” for access to critical AI technologies and risks being cut off without warning. Committee chair Dame Chi Onwurah warned, “The government needs a realistic plan to achieve sovereign capabilities in critical areas or risk having its access cut off at the whim of its partners… Without this, we risk falling even further behind in the global race for science and technology capability, undermining our economic prosperity and national security.” The report urges the government to formulate an explicit AI sovereignty strategy, treating AI as a strategic asset akin to nuclear technology.
The Anthropic Export Ban: A Case Study
The warning gained concrete traction after the Trump administration imposed restrictions on certain frontier AI models built by Anthropic, labeling them a “national security risk.” The export ban barred non‑US users from accessing the company’s most advanced Fable and Mythos models. Anthropic claims these models have already uncovered software vulnerabilities in “every major operating system and every major search engine” in the world, including flaws that have remained undetected for decades. As cyber‑risk expert Joe Hancock of Mishcon de Reya told The Independent, “This episode shows how an advanced AI capability, sold commercially and relied upon operationally, can be withdrawn from every non-US user by unilateral government action at a few hours’ notice, with no transition period.” Although safeguards have since been lifted for most of the controls, the Mythos model remains unavailable to any organization outside the United States.
China’s Counter‑Moves
While the US tightened its grip on Anthropic’s outputs, Chinese authorities were reportedly holding meetings with leading domestic tech firms to consider restricting overseas access to their most capable AI models. Beijing has already blocked Meta’s $2 billion acquisition of the Chinese AI startup Manus, a move designed to keep local talent and intellectual property from flowing to US firms. One Chinese startup even claims to have produced a “cyber nuclear weapon” that matches the capabilities of Anthropic’s Mythos, capable of discovering and exploiting software vulnerabilities in critical infrastructure at an unprecedented scale. These actions mirror the protective logic that drove the 1946 McMahon Act, now applied to the AI domain.
Strategic Rhetoric: AI as a Digital Nuclear Weapon
The analogy between advanced AI and nuclear arms has been echoed by senior US officials. CIA Director John Ratcliffe, speaking at the AWS summit in Washington, declared, “AI, in particular, is a domain in which the CIA must excel, because every algorithmic decision has implications for US strategic advantage and for the national security of all of our people.” He likened the capabilities of cutting‑edge AI models to “digital nuclear weapons.” Such statements underscore the perception that dominance in AI could determine the balance of geopolitical power, just as nuclear supremacy did after World War II.
The Race Toward Superintelligence
Unlike nuclear weapons, which have a relatively clear technological ceiling, the pace of AI development is extraordinarily rapid, making it difficult for nations to catch up once they fall behind. The article notes, “Unlike nuclear weapons, where there is a relatively clear technological endpoint, the extraordinary pace of AI development means other countries may struggle to catch up.” The ultimate goal in this race is the achievement of artificial general intelligence (AGI) or superintelligence—an AI that exceeds human cognitive performance across all domains. At that point, AI systems could self‑improve beyond human control, pursuing goals misaligned with humanity’s interests.
Existential Warnings from Industry Leaders
Leading figures in the AI community have repeatedly warned of the existential threat posed by uncontrolled superintelligence. Sam Altman of OpenAI and Demis Hassabis of Google DeepMind signed a statement from the Center for AI Safety urging governments to treat “the severe risks” of AI with the same seriousness as pandemics and nuclear war. The statement reads, “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.” Their warnings lend weight to the committee’s call for a sovereign AI strategy, emphasizing that national security and economic prosperity hinge on maintaining control over critical AI capabilities.
A Proposed International Treaty
In May, the Machine Intelligence Research Institute (MIRI) released a draft framework modeled on the 1968 Treaty on the Non‑Proliferation of Nuclear Weapons. Titled the “International Agreement to Prevent the Premature Creation of Artificial Superintelligence,” the 50‑page document calls for a binding pact that would prohibit the development of AI systems deemed “catastrophic for humanity” if left unchecked. MIRI explicitly frames the proposal as a bilateral agreement between the US and China, given their substantial lead in AI capabilities. The draft acknowledges, “Bans are not usually the best way to address dangerous technologies, but artificial superintelligence is uniquely dangerous… For a halt to be effective, it must extend to all countries.” To date, no nation has signed the agreement, highlighting the challenges of forging global consensus in this nascent arms race.
Implications for the UK and Global Stability
The convergence of historical precedent, contemporary policy moves, and expert warnings paints a clear picture: the UK’s future security and economic competitiveness may hinge on its ability to develop sovereign AI capabilities rather than relying on the goodwill of allies. As the committee’s report warns, failure to act risks leaving the nation “cut off at the whim of its partners,” echoing the post‑war vulnerability that spurred Britain’s own nuclear program. Simultaneously, the US‑China rivalry over AI threatens to bifurcate the global tech landscape, potentially leaving many nations lagging behind in a race where the finish line is not a weapon but a form of intelligence that could reshape—or even end—human civilization. Policymakers must therefore treat AI sovereignty with the same urgency afforded to nuclear deterrence, balancing openness with the need to protect critical national assets. Only then can the UK hope to navigate the emerging AI arms race without sacrificing its security or prosperity.
https://www.the-independent.com/tech/ai-sovereignty-nuclear-weapon-arms-race-trump-b3011359.html