loading
loading
Ashley’s view Britain wants to pour another £120 million — roughly $160 million — into a government lab that tests the world’s most powerful AI models, an infusion of cash that alone is more than 10 times the annual budget of its US counterpart. That money is part of a new AI package that UK Chancellor of the Exchequer John Healey previewed ahead of his budget presentation to Parliament on Wednesday. The UK government’s decision to devote resources to AI security and research is a response to heightening concerns about AI capabilities and related cybersecurity risks. It contrasts sharply with the approach in the US, where federal efforts to evaluate AI are receiving far less cash and are shared by different parts of government. In addition to the £120 million infusion, the UK also plans to dole out £60 million to a sovereign AI fund to help boost domestic AI companies, and £18 million to boost homegrown chip design. And the government will devote £10 million per year to research on how AI is shaping employment. “We are incredibly well-placed to succeed — Britain has the only trillion-dollar tech sector in the Western world outside of the United States, the UK’s AI Security Institute is a global leader in frontier model testing, and we have the capital market depth to help companies scale and grow,” a senior UK government spokesperson told Semafor. “The new funding commitments in this budget, alongside investments from US firms like OpenAI and Anthropic, demonstrate that the UK is a global destination for AI firms to do business,” the spokesperson said. The UK’s AISI, formed in 2023, has more than 100 staffers, including alumni from labs like OpenAI and Google’s DeepMind, as well as from British intelligence agencies and universities. Healey called the package “a distinctively British approach to AI,” predicting it is unlikely to face trouble in Parliament, where his Labour party holds a large majority. The View From Washington In the US, the Center for Advancing Innovation and Standards for Super Intelligence — recently renamed to reflect President Donald Trump’s choice description for AI — operates on a budget of about $10-$15 million a year within the Commerce Department’s National Institute for Science and Technology. CAISSI, which was formed the same year as its UK counterpart under the Biden administration, has undergone major changes since Trump took office for a second time. It does not currently have a permanent director; the last one, Chris Fall, resigned in July. It was originally called the US AI Safety Institute, but the Trump administration removed “safety” from its name in 2025 to make clear its desire to embrace AI acceleration. “[The UK’s] AISI is an extremely impressive organization, in large part because the UK has invested a great deal in it,” Arthur Tellis, a senior technology fellow at the Institute for Progress, a technology think tank in Washington, told Semafor. “While resource improvements are critical, it is also important that CAISSI [has] empowered technically competent leadership, a direct line to the White House, and a business model that allows it to track and shape frontier AI development,” Tellis said. “Right now, CAISSI is simply too small and distant to do this effectively.” The US entity has ongoing agreements with top AI labs to vet their models before release and describes itself as “industry’s primary point of contact” within the US government. The work that the private sector does with it and its British counterpart is voluntary and not required by law. Some in Washington are clamoring for CAISSI to get more resources. That includes members of Congress, along with White House Office of Science and Technology Director Michael Kratsios. A White House spokesperson told Semafor that President Trump asked for $27 million in funding for CAISSI in his FY2027 budget request, a 145% increase over the current budget. “Our allies and our adversaries recognize that now is the time to build the expertise to understand and oversee frontier AI,” Rep. Lori Trahan, D-Mass., told Semafor. “The United States can’t afford to fall behind. Paltry funding for CAISSI limits its ability to recruit top talent, support the durable guardrails Congress is working to pass, and help America lead on global standards for this fast-moving technology.” The bipartisan bill Trahan proposed alongside Rep. Jay Obernolte, R-Calif., earlier this year would allocate $100 million for CAISSI annually through 2029. Room for Disagreement There are some who will argue that the US government is putting sufficient resources towards AI development — even if that funding isn’t focused on safety and security. The Energy Department’s Genesis Mission, for example, just announced $159 million in awards for work on AI-enabled quantum, science, and energy projects. Trump himself has sought to address safety concerns by convening AI executives at the White House to sign a nonbinding “Accord on Super Intelligence” and establishing a new task force on AI led by Director of National Intelligence Jay Clayton. “American SI models are the most powerful models in the world, and President Trump will ensure that America continues to lead in SI innovation,” a White House official told Semafor in a statement. “CAISSI will continue to play a role in both pre-deployment testing and the Super Intelligence Force. To ensure all components of the SI Force are under one roof, senior CAISSI officials will collaborate with relevant colleagues from the White House,” the official continued. “The SI Force will work together to accelerate innovation, ensure the safe and responsible deployment of new models, and protect the American people.” Notable - The New York Times profiled the UK’s AI Institute in May. - The White House asked OpenAI and Anthropic to refrain from sharing new models with the UK agency before the US government reviewed them, Politico reported in September.
document fetched 2026-10-29 03:03 · canonical https://semafor.com/article/10/27/2026/britains-ai-testing-agency-outguns-us