The Robot Did The Work. Now Who Pays The Tax?

AI is usually discussed through the lens of technology, productivity and innovation. Less attention is given to what may ultimately become one of its most significant consequences: the pressure it places on tax systems that were designed around human labour.

Historically governments have relied heavily on employment-related taxes to fund public services. Individuals earn salaries, pay income tax and contribute to unemployment and social security systems. As businesses increasingly use AI to perform functions previously carried out by employees, a simple but important question emerges: if machines are doing more of the work, where will the tax revenue come from?

The debate around AI has matured considerably over the past few years. Most serious commentators no longer predict widespread unemployment, but rather a gradual shift in how work is performed. Routine and process-driven tasks are increasingly being automated, while human input remains essential in areas requiring judgement, creativity, oversight and accountability.

Even so, the implications for tax systems are difficult to ignore. If a greater share of economic value accrues to technology and capital, and a smaller share flows through traditional payroll structures, governments will inevitably need to rethink tax systems that remain heavily dependent on labour-based revenue.

The fiscal impact of automation extends well beyond a reduction in payroll costs. Employees generate income tax, unemployment contributions and a range of payroll-linked revenues. AI systems do not. At the same time, businesses can generally deduct the cost of software, infrastructure and technology investments, while employing people carries an increasing regulatory and administrative burden. Whether intentional or not, many tax systems currently make automation financially attractive.

Revenue authorities around the world are already considering how tax policy may need to evolve in response. Consumption taxes such as VAT are one possible area of focus, largely because they are less dependent on how many people participate in producing a particular good or service. Corporate taxation is another. As productivity gains increase and more value remains within businesses, governments are likely to pay closer attention to where profits are generated and how they are taxed.

The idea of a "robot tax" has also moved from academic discussion into mainstream policy debate. The phrase itself is probably unhelpful because it conjures up images of governments taxing machines. The real question is whether businesses that materially reduce their workforce through automation should continue to receive the same tax incentives and deductions that were developed for a more labour-intensive economy.

There is no clear answer at this stage, nor is there any global consensus on how automation should be taxed. What seems almost certain, however, is that governments are not going to watch a growing portion of economic activity migrate from people to machines without eventually revisiting how that activity is taxed. Every major shift in the way value is created has ultimately found its way into the tax system, whether through the digital economy, cross-border services or crypto assets. There is little reason to believe AI will be treated differently.

Some policymakers have gone a step further by suggesting that revenue generated from automation-related taxes should be directed towards workforce development and reskilling initiatives. There is a certain logic to the idea that if technology is reshaping the labour market, part of the economic benefit created by that technology should help workers adapt to it.

None of this suggests that businesses should slow their adoption of AI. The commercial advantages are too significant and the competitive pressures too great for that to be a realistic response. It does, however, require a recognition that today's tax treatment of automation should not be assumed to be permanent. As governments grapple with the fiscal consequences of AI, tax policy is likely to become an increasingly important part of the conversation.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.