IA · 5 November 2026 · 3 min read
The £400 Billion Mirage: UK Slashes AI Economic Projections Sourced From Google
In brief: For over two years, UK government officials championed artificial intelligence by citing a projected £400 billion economic windfall. A formal parliamentary response has now revealed that this estimate came entirely from a 2023 study commissioned by Google, while the British government's actual forecasts place the figure at less than a third of that amount.
by Team Mocchi's
For twenty-four months, it served as the cornerstone of Whitehall's technological messaging: artificial intelligence would supposedly deliver a £400 billion ($538 billion) economic boost to the United Kingdom. Used across two successive governments, the figure was frequently cited to calm public anxieties over workforce displacement and justify a light-touch regulatory approach toward leading frontier AI labs. However, that statistical pillar was just dismantled in Parliament.
In a written answer to a parliamentary question, UK AI Minister Kanishka Narayan acknowledged that the government’s own internal economic forecast ranges between £55 billion and £140 billion by 2030, with a median estimate of £120 billion. That represents less than a third of the widely repeated headline number, accompanied by a candid caveat: even this revised figure remains «hugely uncertain».
The origin of the study and Whitehall's echo chamber
As reported by WIRED, the £400 billion forecast did not originate from official treasury modeling or independent civil service research. Instead, it was derived from a 2023 report commissioned directly by Google and conducted by British consulting firm Public First (now PF Global).
Despite having been financed by a commercial tech giant with a direct interest in cloud and AI adoption, the projection was repeated by departments and public agencies for nearly two years. Government ministers had access to the substantially lower internal estimates as early as November 2025, yet state organizations such as Skills England continued publishing the outdated £400 billion projection well into 2026 without clarifying its private sponsorship.
Slower enterprise adoption and the risk of hype
Addressing the wide divergence between the two figures, Jonathan Dupont, head of economics at PF Global, defended the original study by pointing to slower-than-expected deployment across both public administration and commercial sectors: the 2023 model had assumed a rate of organizational integration that simply failed to materialize.
The admission has sparked sharp criticism in Westminster. Dame Chi Onwurah, chair of the Commons Science, Innovation, and Technology Committee, noted that industry hype constitutes a primary barrier to actual public sector digital transformation. Overstating economic dividends, she argued, distorts policymaking and undermines long-term credibility when practical productivity gains take longer to materialize.
Mocchi's take
The British government's statistical retreat underscores a reality we encounter continuously in the field: conflating vendor-funded market studies with realistic operational returns is a recipe for disillusionment. For organizations developing or integrating artificial intelligence, value is never unlocked by top-down macro projections, but through tightly scoped architectures, custom software integrations, and rigorous tracking of unit economics. Evaluating return on investment against specific workflows remains the only reliable safeguard against infrastructure sales hype.