British Government Quietly Adopts AI for Legislation and Policy, Raising Sovereignty Concerns
Key Takeaways
- ▸AI-generated text has already been incorporated into British acts of parliament, with the government keeping this fact confidential until now
- ▸Large language models from US and China-based companies are being used across UK government institutions for drafting legislation, speeches, briefings, and policy decisions
- ▸The wholesale adoption of foreign-controlled foundational models raises serious sovereignty questions, as these systems are designed to persuade rather than calculate, potentially embedding overseas political biases into British governance
Summary
According to an investigation by Financial Times contributor Tim Thorn, the British government has been covertly integrating AI systems into core governmental functions, including the drafting of legislation, with text composed by large language models already appearing in acts of parliament. The revelation came to light when UK officials read a story about the UAE's AI legislation plans with ironic amusement, knowing their own government had already achieved this milestone. The integration extends far beyond lawmaking—AI systems are embedded throughout British institutions, from the Bank of England's interest rate decisions to BBC article editing and education at Oxford University, yet the public remains largely unaware of the scale and speed of this adoption. The investigation, based on interviews with dozens of current and former government officials, technologists, and MPs (most anonymously), reveals a "quiet handing over of control" in the frameworks that underlie government decision-making. A critical concern centers on sovereignty: foundational models powering these systems are built primarily by US and Chinese companies, meaning British governance increasingly relies on foreign-controlled technology designed to be persuasive rather than accurate, potentially embedding the political biases of overseas tech executives into the heart of British democracy.
- Government adoption of AI was a deliberate cabinet-level decision made in November 2024, suggesting systematic rather than ad-hoc implementation across Whitehall
- Even optimistic AI advocates acknowledge the reasonable concern that handing political decision-making power to models built overseas could represent a loss of democratic control
Editorial Opinion
This investigation exposes a troubling gap between the governance of AI adoption and public accountability. While AI's potential utility in government is genuine, the covert integration of foreign-controlled language models into parliamentary lawmaking represents a fundamental democratic concern that transcends typical technology policy debates. The government's apparent view that AI is a solution to institutional dysfunction, combined with the opaque nature of this deployment, suggests policymakers may have prioritized efficiency over sovereignty and transparency—a calculation that deserves far greater public scrutiny and debate before AI becomes further entrenched in Britain's constitutional machinery.
