A recent Guardian investigation has cast doubt on government claims that the UK is attracting billions of pounds in new AI investment. The probe found that much of the announced investment is not new: existing data centres are being rented rather than built, a supercomputer site has not yet started, and promised investments may never materialise. Job creation claims also appear disconnected from reality.
This pattern is not new. In 2009, Gordon Brown announced $1 trillion in financial aid at the G20 summit, a figure that mixed already-promised funds with aspirational pledges. The current AI investment debacle follows a similar playbook, where headline numbers are impressive but the underlying reality is far less so.
Politics runs on announcements, and large numbers make for good news stories. Ministers have a short-term incentive to claim credit for investment, whether or not it is genuinely new. The dominance of the No 10 grid, which demands a daily 'good news' story, entrenches this behaviour.
The UK policymaking establishment also has a love-hate relationship with numbers. Many ministers, civil servants, and media commentators lack scientific or economic literacy, yet they treat precise-sounding figures with excessive confidence rather than scepticism. The assumptions behind the numbers are rarely examined closely, especially when they are politically convenient.
The media plays its part too. Political correspondents often cover high-profile announcements without the subject-matter expertise to unpick the numbers. Rigorous investigative follow-up, as in the AI case, is rare. This distorts public understanding of what policy can achieve and undermines trust in government.
The costs are significant. When announcements fail to translate into reality, scepticism grows not only about the specific claim but about economic policy more broadly. In a country with weak productivity growth and stagnating living standards, this erodes faith in government action to change things.
