In brief
- U.K. Foreign Secretary Yvette Cooper warned that governments need global agreements to manage AI risks.
- Cooper compared the challenge to nuclear safety efforts that followed World War II.
- She called for cooperation between the U.S., China, and other AI powers on safety standards.
Governments risk repeating mistakes made during the dawn of the nuclear age if they wait to create laws governing artificial intelligence, warned UK Foreign Secretary Yvette Cooper.
In an article published on Monday, Cooper said AI offers breakthroughs but also presents new risks as the technology becomes more powerful and widely available.
“Last month, in Shenzhen, China, I saw the extraordinary promise of AI and robotics used for life-saving healthcare,” Cooper wrote. “But the same technologies are also reshaping the future of warfare, crime and social cohesion in alarming ways.”
Cooper said managing the risks posed by AI may become “the greatest security challenge of the next decade” and argued governments need international agreements around frontier technology before a crisis forces action.
She compared the current race to develop AI systems with the early nuclear arms race, saying global safety agreements emerged only after countries witnessed the devastation caused by atomic weapons.
“On nuclear, international agreement came only after the world saw the terrifying power of the new technology at Hiroshima,” Cooper wrote. “We cannot afford to wait for an AI equivalent of Hiroshima before we act.”
Cooper called on Britain to use its diplomatic influence to bring together the United States, China, and other major AI powers to establish shared safety principles and standards, pointing to the 2023 AI Safety Summit at Bletchley Park, where global leaders from 29 countries and the European Union met to discuss the emerging risk of AI. Cooper called this an example of the UK’s ability to “rally the world on AI security.”
Cooper’s warning comes after months of escalating concerns over how governments should oversee increasingly powerful AI systems.
In May, the U.K.’s AI Security Institute warned that rapid gains in AI cybersecurity abilities after OpenAI’s GPT-5.5 became the second model to complete a simulated cyberattack without human assistance, following Anthropic’s Claude Mythos Preview.
Days later, the International Monetary Fund warned that AI could “amplify” cyberattacks against the global financial system by lowering the skill needed to exploit vulnerabilities, urging policymakers to treat cybersecurity as a financial stability issue rather than a purely technical problem.
In June, President Donald Trump signed an executive order creating a voluntary framework for reviewing advanced AI models before release, expanding AI cybersecurity programs, and directing agencies to evaluate potential national security risks from frontier AI models.
That same month, calls for stricter rules have also come from within the AI industry, with Anthropic CEO Dario Amodei arguing that transparency requirements are no longer enough and calling for mandatory third-party testing of frontier models. This demand was followed by the U.S. government ordering Anthropic to restrict access to Claude Fable 5 and Mythos 5 over national security concerns before lifting the order in July.