The prospect of income tax becoming redundant within five years is no longer being framed as a distant theory. Tom Blomfield, the Monzo founder, has argued that artificial intelligence could push governments toward taxing computing power instead of human labour. The claim lands at a moment when AI is already changing the jobs market, and when the UK’s biggest revenue streams still depend heavily on wages. The timing makes the warning less like a prediction about the future and more like a challenge to the present.
Why this matters now for income tax and public revenue
Blomfield’s argument is simple but disruptive: if AI keeps replacing work, the tax base built around employment weakens. He said he does not think governments will continue to tax human labour in the same way, and instead expects systems such as data centres to become a target for revenue collection. That idea directly collides with the current structure of public finance, where income tax and National Insurance contributions remain the two biggest sources of government revenue in the UK, making up 42 per cent of the total.
The concern is not just about how money is raised, but whether the existing model can keep pace with the speed of automation. Blomfield said AI agents are already performing narrow tasks better than humans in some cases, including work that has traditionally depended on professional expertise. He suggested that by the end of 2026, these tools could become generalisable, a shift that would widen the pressure on jobs beyond the most obvious sectors.
What lies beneath the income tax debate
At the center of the debate is a basic fiscal problem: the state needs a reliable way to fund services if fewer people are paid wages. Blomfield argued that professions such as tax accounting could require almost no humans in the near future. If that proves accurate, then a tax system built mainly around salaries becomes harder to defend, especially if automation expands faster than new forms of employment emerge.
The numbers in the wider discussion show why policymakers are being pushed to think differently. Taxes on capital gains, property sales, stamp duty and inheritance tax make up just four per cent of government revenue, far smaller than the share tied to income tax and National Insurance. That gap highlights how dependent the system remains on labour, even as the labour market changes.
There are already signs of strain. Jobs site Adzuna said adverts for entry-level roles were down 35 per cent compared with November 2022, when ChatGPT launched. Morgan Stanley has also warned that Britain could be especially exposed to an AI-led jobs shock because of its heavy reliance on professional services, which accounted for 81 per cent of economic output last year. Together, those figures suggest the pressure on income tax is not only theoretical.
Expert perspectives on tax, AI and the state
Blomfield, speaking on The Rest is Money podcast, said: “I don’t think we’ll tax human labour, we’ll tax compute, [meaning systems like] data centres, and then we will use the proceeds to pay for government. ” He also said that AI systems are already beating humans in narrow domains and described them as “very narrow geniuses” for now.
An episode synopsis for the podcast posed the central policy question directly: if AI replaces all or most jobs, how can governments replace lost taxes, and who will pay for public services? That framing captures the scale of the issue. Last week, OpenAI suggested policymakers may need to shift toward taxing capital, corporate profits and long-term AI-driven returns, potentially including a robot tax on automated labour.
The analysis here is that the debate is moving from productivity gains to fiscal survival. If AI cuts labour demand while boosting output, governments may need to decide whether to tax the machines, the compute, or the profits they create. Each option has political and practical obstacles.
Regional and global impact of an AI tax shift
For the UK, the implications go beyond one tax category. A move away from labour-based revenue would reshape how the Treasury thinks about work, business investment and economic fairness. It could also increase pressure on countries with service-heavy economies, where professional jobs are especially exposed to automation. That is why the warning over income tax resonates far beyond one podcast appearance.
Globally, the idea of taxing compute rather than people suggests a new policy frontier. But it also raises a harder question: can governments redesign revenue systems quickly enough to avoid a gap between disappearing jobs and disappearing taxes? If AI keeps advancing at the pace Blomfield described, that question may arrive sooner than many policymakers expect.
For now, the core issue is not whether income tax disappears overnight, but whether governments can adapt before the old model stops working.
