Universal Basic Income in the Age of Artificial Intelligence

Saturday 18 July 2026

The idea of a universal basic income (UBI) has moved from the fringes of political philosophy to the centre of economic debate. For decades it was regarded as an interesting but impractical proposal advanced by academics, futurists and a handful of social reformers. Today it is discussed in ministries of finance, international organisations and boardrooms alike. The reason is simple: the accelerating development of artificial intelligence has revived an old question. What happens if technological progress reduces the demand for human labour faster than new occupations can be created?

Universal basic income is a deceptively simple concept. Every citizen receives a regular payment from the state, regardless of employment status, wealth or personal circumstances. Unlike traditional welfare benefits, it is unconditional. One receives it whether one works or not. The purpose is not merely to alleviate poverty but to establish a minimum economic floor beneath every member of society.

The concept is far older than many imagine. Variants can be found in the writings of Enlightenment thinkers and nineteenth-century social reformers. Yet it was only in the twenty-first century that governments began conducting substantial experiments to determine whether the idea might function in practice.

The most famous experiment took place in Finland between 2017 and 2018. Two thousand unemployed individuals received a monthly payment of €560 without conditions.

Researchers wished to determine whether removing welfare bureaucracy and work disincentives would encourage recipients to seek employment. The results proved intriguing. Employment effects were modest, with little evidence that recipients worked substantially more than those receiving conventional benefits. However recipients reported higher levels of wellbeing, lower stress and greater economic security. They expressed more confidence in their futures and experienced less psychological strain.

The Finnish experiment therefore demonstrated both the strengths and limitations of UBI. It appeared capable of improving quality of life, but it did not generate the dramatic employment gains that some advocates had predicted. Critics seized upon the latter finding, while supporters emphasised the former.

Finland was not alone. Smaller experiments have occurred in Canada, the United States and elsewhere. One widely discussed American pilot, supported by organisations associated with artificial intelligence entrepreneur Sam Altman, provided regular cash payments to participants over several years. The findings suggested improvements in financial resilience and personal autonomy, although the effects on labour participation were again more nuanced than either advocates or opponents had expected.

Historically, sceptics of UBI have always asked the same question: if people are paid regardless of whether they work, why would they continue working? For much of modern economic history, that objection carried considerable force. Labour was essential because there was no realistic prospect of replacing large numbers of workers with machines.

Artificial intelligence has altered that calculation.

The contemporary debate differs fundamentally from earlier discussions because it is no longer focused solely upon poverty reduction or welfare reform. Instead it concerns the future structure of labour markets themselves.

The International Monetary Fund estimates that artificial intelligence could affect approximately 40 per cent of jobs worldwide and as many as 60 per cent in advanced economies. Some occupations will be complemented by artificial intelligence, making workers more productive. Others may be partially or wholly replaced.

What makes the present technological revolution unusual is that it threatens not only routine manual work but also many forms of white-collar employment. Previous waves of automation primarily affected factory workers, agricultural labourers and clerical staff. Artificial intelligence is increasingly capable of performing tasks once thought uniquely human: drafting documents, analysing data, producing computer code, conducting legal research and generating creative content.

The consequences are already becoming visible. Research suggests that regions and occupations with greater exposure to artificial intelligence have experienced measurable changes in employment patterns and hiring behaviour. Entry-level positions appear especially vulnerable because many of the routine tasks assigned to junior employees can now be performed by sophisticated software.

Yet predictions of mass unemployment should be treated cautiously. Every major technological revolution has generated fears that proved exaggerated. The mechanisation of agriculture destroyed millions of farming jobs, but industrial employment emerged. Industrial automation reduced manufacturing employment, yet service industries expanded dramatically. Recent evidence suggests that artificial intelligence is simultaneously eliminating some occupations while creating demand in others. Employment patterns are being reorganised rather than simply erased.

This is where the debate becomes particularly complex.

If artificial intelligence ultimately creates as many jobs as it destroys, then universal basic income may be unnecessary. Existing welfare systems, combined with retraining programmes and labour-market reforms, may suffice.

However if artificial intelligence produces a more concentrated economy in which fewer workers generate vastly greater economic output, then a different challenge emerges. Wealth creation may continue growing even as labour’s share of national income declines. In such circumstances, traditional social insurance systems linked to employment become increasingly difficult to sustain.

The real significance of universal basic income may therefore lie not in combating unemployment but in addressing economic insecurity.

Modern labour markets are becoming increasingly fragmented. Short-term contracts, freelance work, platform-based employment and intermittent careers are replacing the stable lifetime employment model that characterised much of the twentieth century. Artificial intelligence may accelerate these trends. Individuals may move repeatedly between periods of employment, retraining and entrepreneurial activity.

Under such conditions, an unconditional income floor becomes easier to understand. Rather than serving as a substitute for work, it functions as a stabiliser within a more volatile economy.

There are nevertheless formidable objections.

The first is cost. A genuinely universal basic income requires enormous public expenditure. Even wealthy countries would need substantial tax increases, spending reductions elsewhere or significant borrowing. The mathematics quickly become daunting. This is why many economists who sympathise with the objectives of UBI remain sceptical about its affordability.

The second objection concerns incentives. Although most experiments have not demonstrated dramatic reductions in work effort, large-scale national programmes might produce different results. Human behaviour often changes when temporary experiments become permanent institutions.

The third objection is philosophical. Critics argue that income should be linked to contribution. A society in which millions receive income without participating in productive activity may experience declining social cohesion. Employment provides more than money; it provides structure, purpose, status and community. No cash payment can fully replace those functions.

This criticism deserves serious consideration. Much contemporary discussion of artificial intelligence assumes that work is merely a means of obtaining income. For many people it is also a source of identity. The challenge posed by artificial intelligence is therefore cultural as much as economic.

Perhaps the most likely future lies between the extremes.

Rather than implementing a full universal basic income, governments may gradually expand forms of guaranteed income support, simplify welfare systems and create more flexible social safety nets. They may combine these measures with investments in education, lifelong retraining and public services. Such reforms would address many of the same concerns that motivate advocates of UBI without requiring a complete transformation of fiscal policy.

The experiments conducted thus far suggest that universal basic income is neither a utopian solution nor an economic catastrophe. It appears capable of improving wellbeing and reducing insecurity, but it is not a magic formula for employment growth. Its greatest significance may be as a lens through which societies examine deeper questions about technology, work and human dignity.

Artificial intelligence has not yet rendered human labour obsolete. Indeed it may never do so. Yet it has forced governments, economists and citizens to confront a possibility that previous generations could safely ignore: a future in which economic prosperity becomes increasingly detached from human employment.

If that future arrives, the debate over universal basic income will cease to be a theoretical exercise. It will become one of the defining political questions of the twenty-first century. The challenge will not merely be how wealth is created, but how it is distributed in a society where intelligent machines increasingly share the burden of production. Universal basic income is not really about welfare at all. It is about redefining the relationship between citizenship, work and economic security in an age of artificial intelligence.