UK Foreign Secretary Warns of 'AI Hiroshima' Without Global Safety Pacts

Cooper highlighted the dual nature of AI, acknowledging its potential for breakthroughs in areas like healthcare, as she witnessed in Shenzhen, China, but also emphasizing its alarming implications for warfare, crime, and social cohesion. She described managing AI risks as potentially "the greatest security challenge of the next decade" and argued for international agreements on frontier technology to be established proactively, rather than waiting for a crisis.

She invoked the history of nuclear safety, noting that global agreements only emerged after the devastation of Hiroshima. "We cannot afford to wait for an AI equivalent of Hiroshima before we act," Cooper wrote. She urged Britain to leverage its diplomatic influence to unite the United States, China, and other major AI powers in developing shared safety principles and standards, citing the 2023 AI Safety Summit at Bletchley Park as an example of the UK's capability in this area.

Cooper's warning follows a period of escalating concerns regarding AI oversight. In May, the UK's AI Security Institute reported significant advancements in AI's cybersecurity capabilities, with models completing simulated cyberattacks. Shortly after, the International Monetary Fund cautioned that AI could amplify cyberattacks against the global financial system by lowering the skill threshold for exploiting vulnerabilities.

In June, President Donald Trump signed an executive order aimed at reviewing advanced AI models and expanding cybersecurity programs. Concurrently, industry leaders like Anthropic CEO Dario Amodei called for mandatory third-party testing of frontier AI models, a sentiment echoed by U.S. government actions that temporarily restricted Anthropic's access to certain models over national security concerns before lifting the order.