Artificial intelligence and robotics could set off an economic transformation whose social consequences we can still only begin to imagine. While Europe struggles to remain competitive, shoulder the economic costs of war and finance its welfare states, much of its political debate remains trapped in the ideological divisions of the last century. Yet the defining question of the coming decades is unlikely to be how to provide an unconditional income to workers whose jobs have been made redundant by technology. It is whether we can build an economy and a society in which human work finds new purpose and new opportunities. The twentieth century created the welfare state. The twenty-first must now work out how to preserve it while the very meaning of work is changing.
Author: Szilárd Szélpál
The Lamplighters Were Not Given a Basic Income
In nineteenth-century European cities, keeping the streets lit was still very much a human job. Every evening, lamplighters walked the streets lighting gas lamps, returning in the morning to put them out. It was an established occupation, with its own routines, place in society and economic importance. Then came electricity, power grids and automated street lighting, and the need for lamplighters gradually disappeared. Not because they were bad at their jobs, or because society had suddenly lost interest in them, but because technological progress had made the work they depended on for a living unnecessary.
Electricity, of course, also created new industries, professions and economic opportunities. Electricians, engineers, power station workers, network installers and maintenance technicians were suddenly in demand. Technological change increased productivity and helped raise living standards. But the jobs created by the new economy did not necessarily appear in the same places, at the same time, or for the same people who had lost their previous livelihoods. A middle-aged lamplighter did not become an electrical engineer overnight. Nor does today’s office worker automatically become an AI developer simply because an algorithm can now perform much of their work.
The economic upheaval of the second half of the nineteenth century brought considerable social tensions across Europe. Industrialisation, changes in agricultural production, land shortages, population growth and economic inequality drove millions of people to leave their homes. Nearly twelve million immigrants arrived in the United States between 1870 and 1900, followed by more than fifteen million over the next fifteen years. Naturally, they were not all fleeing occupations made obsolete by technology. Migration was driven by a far more complex combination of economic, demographic, social and political forces. Nevertheless, the period offers a useful reminder that economic progress does not automatically mean a smooth transition for the societies affected.
The institutions of the modern welfare state developed partly in response to precisely these kinds of social upheavals. Social insurance, workers’ rights, the expansion of public education and state welfare policies reflected a growing recognition that industrial society could not rely entirely on the market’s ability to adjust on its own. Conservative reformers, liberal politicians, Christian democratic movements and social democrats all contributed to the development of the welfare state. Labour movements, in particular, became a major political force in securing social protection for working people.
Now, more than a century later, we face another technological turning point that could fundamentally reshape the world of work. This time, however, we already have institutions, economic resources and scientific tools that could allow us to prepare for the changes rather than simply deal with their consequences afterwards.
History, then, has a few lessons to offer. It would be a shame to wait once again until the social consequences force us to pay attention.

Artificial Intelligence Does Not Need a Work Permit
One of the distinctive challenges posed by artificial intelligence is that it no longer affects only physical labour and routine tasks. Earlier waves of automation mainly transformed industrial production, agriculture and certain relatively simple services. Generative AI, by contrast, can increasingly be used in work that we once associated with the relative security of highly educated white-collar professions.
According to research published by the International Labour Organization in 2025, around 24 per cent of jobs worldwide are exposed to generative AI to some degree. In high-income countries, the figure is roughly one-third. Administrative, accounting, data-processing and secretarial roles are among the most exposed, but automation is also becoming a real possibility in a growing number of financial, IT and analytical occupations.
This does not mean that all these jobs will automatically disappear. The ILO itself considers the transformation of occupations more likely than their immediate, large-scale elimination. What is difficult to dispute, however, is that AI could change demand for workers, wage structures, employers’ hiring decisions and the distribution of income generated by economic activity.
An especially interesting finding comes from the ILO’s 2026 analysis, which suggests that women may face greater risks from automation because they are overrepresented in many administrative and service occupations likely to be affected by the technology. So while much of Europe’s political debate continues to revolve around cultural conflicts over gender identity, technological progress may be reshaping a very real gender inequality in the labour market. It is a rather less glamorous problem than drafting another ideological declaration, but a considerably more serious policy challenge.
For much of society, the immediate prospect is not one of economic opportunity. It is the uncertainty of whether their existing skills will still have any market value in ten or fifteen years. Will they keep their jobs? Will they still be able to provide a decent income for their families? Will they have the chance to learn a new profession? And what happens if technology advances faster than their ability to adapt?
These are not abstract questions. For a worker, the technological revolution is not a productivity curve in a corporate presentation. It is a very personal threat to their livelihood.
The job of politics should be to recognise the difference.
Europe Wants to Compete, but Has Other Bills to Pay
Europe’s economic position makes the challenge of technological adaptation particularly serious. The United States and China enjoy substantial advantages in several critical technological fields, from AI infrastructure, cloud services and semiconductor manufacturing to the space industry. Europe, meanwhile, is still trying to work out how to become a major player in the technological and economic competition of the coming decades.
Nor is the continent attempting to catch up in a vacuum. The war in Ukraine, changes in energy markets, rising defence expenditure and the growing costs of ageing societies are all placing pressure on public budgets. The war also came on top of inflationary and energy-market tensions that had already begun to emerge, leaving European economic policymakers with even less room for manoeuvre.
According to figures updated by the European Parliamentary Research Service in October 2026, the European Union and its member states have mobilised approximately €224.5 billion in financial, military and humanitarian support for Ukraine since February 2022, a substantial share of it through loan arrangements. At the 2025 NATO summit in The Hague, heads of state and government also committed to spending five per cent of GDP on defence and security-related purposes by 2035, with at least 3.5 per cent allocated to core military expenditure.
These commitments present European governments with difficult budgetary choices, at precisely the moment when much greater investment is needed to maintain the continent’s technological competitiveness. In July 2026, for example, the European Commission launched a new AI gigafactory programme involving up to €10 billion in EU and national funding, with the expectation of attracting at least another €20 billion in private investment. An important initiative, certainly. But the sheer difference in scale between these sums gives us something to think about.
Of course, support for Ukraine is not one large pot of money that could simply be spent elsewhere. Loans, military equipment and budgetary assistance are financially different from industrial investment. Supporting Ukraine’s resistance also serves European security interests, which cannot simply be crossed out of a budget spreadsheet.
None of this, however, makes the question of opportunity costs irrelevant. If Europe can mobilise hundreds of billions of euros for geopolitical objectives, it is reasonable to ask whether it can show comparable political determination when financing its own technological and economic future. Funding a war may be a security necessity, but no country or economic community can build lasting prosperity solely by responding to necessities.
Europe therefore needs to do more than simply spend more money. It must also direct its resources towards sectors capable of creating new industries, high-value jobs, technological independence and sustainable tax revenues. Artificial intelligence, robotics, the commercial space industry, semiconductor technology, energy systems and dual-use innovation all offer such opportunities.
Because Europe cannot keep redistributing prosperity if it plays an ever smaller part in creating it.
A Welfare State Still Built Around Yesterday’s Worker
The European welfare state largely developed around an economic model in which people have jobs, earn wages and contribute to public finances through income taxes and social security payments. These revenues help fund pensions, healthcare, unemployment benefits and a wide range of social services.
Naturally, governments have other sources of revenue as well. Consumption taxes, corporate income taxes and various taxes on wealth and property also play important roles. Nevertheless, income from work still accounts for a substantial share of public revenue in many member states.
But what happens when an increasing share of economic value is created by automated systems, while some companies can produce the same output with considerably fewer employees?
Productivity may rise, profits may increase, and employment-related social contributions may fall. Of course, automation can also create new jobs, higher wages and additional tax revenue. The economy is not a closed system in which every lost job represents a permanent loss. Even so, the question remains: are our tax systems adapting adequately to an economy in which technological capital plays an increasingly important role in creating value?
The answer is hardly to introduce a special tax on every robot and AI system. That could easily make precisely the technological investments Europe most needs in global competition more expensive. This is one reason why the IMF warned against a straightforward robot tax in its 2024 analysis, while also arguing that the taxation of capital income and corporate profits deserves reconsideration.
The real objective should not be to tax technology itself, but to ensure that a fair share of the additional value generated by automation contributes to public finances. This could involve modernising corporate taxation, reviewing tax incentives that unnecessarily favour automation, taxing capital income more effectively and reducing international tax avoidance.
Any additional or newly available public resources should then be used not merely to compensate people indefinitely for lost earnings, but to develop the skills and services needed in the new economic environment.
The distinction matters. The state should not be artificially keeping obsolete jobs alive. It should be creating new opportunities for the people who once held them.
The Comfortable Promise of Basic Income and the New Meaning of Work
Universal basic income has been a recurring political proposal in response to the possibility of technological unemployment. At first glance, it is an appealing idea: if machines can perform an increasing share of human work, the state should provide everyone with a fixed, unconditional income.
The social concern behind the proposal is understandable. If technological progress makes employment income less secure, guaranteeing a minimum level of financial security could reduce people’s vulnerability.
But earning a living and participating in society are not the same thing. A regular income does not, by itself, provide professional pride, a sense of belonging, opportunities for personal development or the feeling that one’s work is useful to others.
International experiments with basic income have also produced mixed labour-market results. In Finland, the effects on employment were relatively modest, while participants reported improved well-being. An American programme examining unconditional cash payments, meanwhile, recorded reductions in labour-market participation and working hours. These findings do not prove that every basic income scheme necessarily discourages employment. They do show, however, that providing income security alone does not solve the problem of adapting to technological change.
The welfare system of the twenty-first century should therefore set itself a far more ambitious goal. It should not merely ensure that people have some income even when they are not working. It should help them develop, repeatedly if necessary, the ability to contribute value as the economy changes around them.
That would require fundamental changes to education, training and employment systems. Governments, businesses and educational institutions should work together to anticipate which occupations are likely to face substantial automation and where new employment opportunities may emerge.
Retraining should not begin only after someone becomes unemployed, when they have already lost part of their income, their professional confidence has been shaken and their family’s financial security may be at risk. Opportunities to acquire new skills should be available while people are still employed, as part of a planned and continuous process.
Technological transition does not have to mean social decline. But avoiding that outcome requires policies that do more than rescue people after they have lost out. They must prepare people in advance, so that they can take part in change rather than simply suffer its consequences.
The Opportunity for a New Service-Based Society
Much of the human labour freed up by technological progress could be put to use in sectors where Europe already faces persistent staff shortages, limited capacity or declining service standards.
Healthcare, elderly care, education, rehabilitation, family support services and community infrastructure are all areas where demand for workers may continue to grow over the coming decades. For Europe’s ageing societies, developing services that require personal care, coordination, organisation and human relationships will be especially important.
Someone whose administrative job has disappeared because of automation might, with appropriate training, find work as a healthcare coordinator, education administrator, patient care coordinator, digital services adviser or community services organiser. Not everyone is suited to every role, of course, and a short course cannot replace years of professional training. Opportunities must be matched to individual abilities and genuine labour-market needs.
The point is that society should not treat automation solely as a loss of employment income. It should also see it as a release of human capacity that could be used to create new services.
This would also require the new jobs to offer more than lower wages and less secure employment than the positions they replace. The transition can succeed only if workers have access to real alternatives that provide a decent living and enjoy social recognition.
Such a model would also fundamentally change the role of the state. Welfare would no longer consist primarily of income transfers, but also of organising services, training and employment opportunities that could improve society’s productivity and quality of life over time.
After all, the social purpose of technological progress is not necessarily to have fewer people doing the same work. It can also be to allow people to create value elsewhere.
Solidarity Has to Be Sustainable Too
A new social contract of this kind cannot be built solely on expanding benefits and entitlements. For welfare systems to remain sustainable, as many working-age people as possible must participate in creating economic and social value.
Social solidarity cannot mean that the state should have no expectations of those receiving benefits. Work must pay, eligibility must be checked, and abuse must carry consequences. At the same time, we need to distinguish clearly between people who cannot work, those who want to work but lack suitable opportunities, and those who are demonstrably claiming benefits to which they are not entitled.
This distinction will become especially important as technological change pushes more workers into periods of uncertainty. Society cannot afford to turn people who need help adapting into permanent clients of the welfare system. Nor can it abandon those who genuinely need support under the pretext of fighting abuse.
The same consistency is needed when addressing the economic and social questions surrounding migration. Many immigrants who work legally make valuable contributions to European economies and public finances. Long-term inactivity, unsuccessful integration and abuse of welfare systems can, however, present genuine policy problems. The response should be based on verifiable eligibility, ability to work and individual conduct, rather than the collective stigmatisation of particular social groups.
The state must protect both those who genuinely need help and the public legitimacy of the system itself. If working people come to believe that the balance between contributing to public finances and receiving benefits has been permanently lost, sooner or later the institutions of solidarity will lose public support.
The long-term survival of the welfare state therefore requires more than redistribution. It must also encourage social participation, individual responsibility and the creation of value.
Tax the Rich! An Old Recipe for Another Set of Budget Problems
When parts of progressive politics run short of ideas for economic renewal, an old and well-tested slogan tends to reappear: tax the rich! The proposal certainly has political appeal. It is easy to understand, invokes moral fairness and reassures most voters that the growing costs of the welfare state will, this time, be paid by someone else. How much revenue the measure will actually raise, what it will do to investment, how capital might move, or how taxpayers might change their behaviour are rather less suitable subjects for campaign posters. Tax policy, however, has an inconvenient habit: actual revenue does not necessarily follow the logic of political promises.
There is plenty to learn from international experience. According to a comparative OECD study published in 2018, twelve member countries still levied recurrent taxes on individuals’ net wealth in 1990. By 2017, only four did. Austria, Denmark, Germany, Finland, the Netherlands and Sweden had all abolished their previous wealth taxes, while France largely restricted its version to real estate in 2018. The reasons varied, but recurring problems included the difficulty of valuing assets, administrative costs, tax avoidance, the risk of capital flight and revenues that fell short of expectations. Such taxes have not disappeared altogether, however. Norway, Spain and Switzerland still use them, a reminder that not every wealth tax produces the same results.
Nor is this merely a historical debate. According to the OECD’s September 2026 report on financing social protection, recurrent net wealth taxes accounted for just 0.4 per cent of total tax revenues across OECD and EU countries in 2024. A European Commission study published the same year also found that the wealth taxes examined generally did not provide a major source of revenue, and that their effectiveness depended heavily on their design, exemptions and taxpayer behaviour. None of this proves that wealthier groups could not contribute more to public finances. It does, however, show rather clearly that taxing wealth and collecting substantial, reliable revenue are not necessarily the same thing.
The latest chapter in this debate is unfolding in California, where voters are due to decide on 3 November 2026 whether to introduce a one-off five per cent tax on individuals with wealth exceeding one billion dollars. Supporters of the initiative, known as Proposition 40, estimate that it could raise around $100 billion, with 90 per cent earmarked for healthcare. The objective sounds admirable, and the political message is wonderfully straightforward: the wealthiest should help pay for public services. The people expected to pay, however, do not always share the assumption that where they live and pay taxes is an unchangeable fact of life, regardless of political decisions.
In an October 2026 paper published through the US National Bureau of Economic Research, Benjamin Jaros and Joshua Rauh found that the publicly known departures of seven billionaires previously associated with California already affected nearly 30 per cent of the potential tax base. Their model estimates that the one-off tax would raise around $30 billion. Once the future personal income tax revenues associated with departing taxpayers are taken into account, however, they calculate a long-term net present value loss of $38.9 billion. This is a serious warning, although it remains a forecast rather than a final budgetary outcome. California’s official Legislative Analyst’s Office offers a more favourable assessment, projecting a temporary revenue increase of tens of billions of dollars and an ongoing annual personal income tax loss of less than $1 billion. The gap between these estimates is itself a useful illustration of how far we are from being able to establish who will pay, how much and for how long on the strength of a catchy tax slogan.
One of the most interesting figures in the California debate is Meta founder Mark Zuckerberg. The proposed wealth tax could leave him facing a liability worth billions of dollars, even though much of his wealth is held in company shares rather than in a bank account conveniently set aside for paying taxes. According to corporate documents released by Meta on 21 October 2026, Zuckerberg had previously considered moving his residence outside California, while the company agreed with him on the creation of a new class of non-voting shares. This could allow him to raise the liquidity needed to meet a potential tax obligation while retaining voting control over the company. The case illustrates one of the central difficulties of modern wealth taxation: someone can be extraordinarily wealthy on paper, while raising the cash needed to pay a tax may have consequences for corporate governance and financing decisions.
None of this means that taxing large fortunes is inherently pointless, or that capital income should enjoy some special exemption. Wealth inequality and the concentration of income from economic activity are genuine social policy concerns. But a tax system must be economically workable as well as fair. There is a significant difference between taxing a company’s actual profits, realised capital income and a stock of wealth partly made up of unrealised gains. Each creates different economic incentives, affects investment differently and carries different risks for public finances.
And this brings us back to artificial intelligence. If Europe genuinely wants a fair share of the productivity gains from automation to help sustain its welfare states, it does not necessarily need to invent new taxes on the wealth of business owners. It would make more sense to develop a tax system that effectively captures actual corporate profits and capital income, reduces the disproportionate burden on labour, limits artificial profit shifting and avoids unnecessarily penalising innovation. Any additional revenue should then be used not simply to create more benefit entitlements, but to build training, employment and service systems through which people can genuinely share in the benefits of technological progress.
Social justice, after all, is not achieved simply by identifying a group from which politicians would like to collect more money. Someone also needs to work out how to create more economic value, how to finance public services sustainably and how to ensure that taxation does not undermine the very investments expected to generate future revenues. Especially in an economy where capital and technology are becoming increasingly indifferent to national borders.
For politicians, of course, the old slogan is far more convenient. It is much easier to promise an answer to the question of how to tax the rich than to explain how an ageing continent struggling to keep up technologically should rebuild the financial foundations of its welfare system. But if we are serious about creating a social contract for the twenty-first century, perhaps it is time to take an interest in the twenty-first-century economy as well. Reopening the tax debates of the last century does not automatically count as progress, even when the campaign comes with better graphics and somewhat louder promises of fairness.
A Political Class Still Fighting Yesterday’s Battles
The technological transformation now under way raises some uncomfortable questions for traditional political ideologies. Historically, the left sought to strengthen social protection for working people. Much of the economic tradition of the right, meanwhile, emphasised entrepreneurship, private property, investment and market competition. In the new technological environment, neither approach is sufficient on its own.
Social redistribution without adequate economic performance eventually becomes impossible to finance. Technological progress without social adaptation, on the other hand, can create inequalities and insecurity that undermine public acceptance of the economic system itself.
What politics needs in this situation is not another search for new slogans. It needs a better understanding of how these issues fit together.
Parts of the left continue to frame social justice primarily in terms of equal rights and identity politics. Yet as technology advances, economic security, employability, retraining and the financing of redistribution may become increasingly important questions of social justice. Right-wing populist movements, meanwhile, are often successful at recognising public dissatisfaction and the desire for security. That does not necessarily mean they have workable economic policies to offer.
According to a 2026 Pew Research Center analysis, the rise of right-wing populism in Europe reflects a combination of economic grievances, cultural conflicts and distrust of political elites. In other words, loudly identifying a social problem is not enough. Political credibility now requires institutions, economic incentives and practical programmes capable of addressing it.
People also expect more from politics than higher incomes alone. They want security, predictability, decent public services, affordable housing and a future they can plan for. According to the spring 2026 Eurobarometer, 52 per cent of Europeans regard the cost of living as their main personal concern. Uncertainty about the technological future may make these expectations even more pressing.
It is therefore rather peculiar to watch politicians repeatedly reaffirm their ideological identities in response to social challenges that primarily require economic restructuring and institutional innovation.
Justice does not become a reality simply because politicians mention it often enough. It becomes real when people actually have the means to benefit from progress.
On the Threshold of a New Social Contract
One of the most important tasks facing European politics in the coming decades may be to establish a new relationship between technological progress, economic value creation, taxation and social security.
That would require faster development of European AI and robotics, a fair contribution to public finances from the profits generated by new technologies, retraining opportunities for workers before their jobs disappear, and greater employment capacity in socially important services. Some welfare measures should be turned into investments that do more than soften economic insecurity: they should also reduce the likelihood of it arising in the first place.
This does not mean dismantling the social safety net. Quite the opposite. Social protection needs to be adapted to an economic model in which occupations change more quickly, skills can lose market value in a shorter time, and working lives become increasingly less linear.
The modern welfare state should therefore promise more than ensuring that nobody is left without an income. It should also promise that nobody becomes unnecessary simply because a machine can perform their former job more efficiently.
That could provide the foundation for a genuinely twenty-first-century social contract. Not necessarily left-wing or right-wing, but based on the recognition that technological progress and social well-being are not separate objectives. They are mutually dependent parts of the same economic system.
The historical opportunity is there. The only question is whether Europe’s political systems can recognise it in time, or whether they will once again wait until the costs of social adaptation can no longer be hidden behind budget reports and political declarations.
The lamplighters of the nineteenth century at least knew that electricity had taken their jobs. We expect rather more from workers in the twenty-first century: they must understand the economic consequences of artificial intelligence, retrain, adapt to a rapidly changing labour market and, preferably, continue financing the welfare state while they are at it.
Perhaps it is time to stop expecting workers to do all the adapting. Because if society must prepare for the age of artificial intelligence, it would not be entirely unreasonable to expect politics to show a little intelligence of its own.
Sources
- International Labour Organization – Generative AI at Work: What It Means for Jobs in Europe and Beyond (2025)
- International Labour Organization – New ILO Data Confirm Women Face Higher Workplace Risks from Generative AI than Men (2026)
- International Monetary Fund – Broadening the Gains from Generative AI: The Role of Fiscal Policies (2024)
- OECD – Skills in the AI Age (2026)
- OECD – Skills Outlook 2025 (2025)
- European Parliamentary Research Service – State of Play: EU Support to Ukraine (2026)
- NATO – The Hague Summit Declaration (2025)
- European Commission – EU Launches AI Gigafactories Call (2026)
- European Commission – Standard Eurobarometer 105 (2026)
- Pew Research Center – Right-Wing Populism in the Decade Since Brexit (2026)
- Library of Congress – Immigration to the United States, 1851–1900
- OECD – The Role and Design of Net Wealth Taxes in the OECD (2018)
- OECD – Financing of Social Protection (2026)
- European Commission – Study on Wealth Taxation, Including Net Wealth, Capital and Exit Taxes (2026)
- Jaros, Benjamin & Rauh, Joshua – The Net Present Value of the California Billionaire Tax Act, NBER Working Paper No. 35899 (2026)
- California Legislative Analyst’s Office – Proposition 40: Analysis by the Legislative Analyst (2026)

Szilárd Szélpál served as an environmental expert in the European Parliament from 2014, where he utilized his expertise to influence policy-making and promote sustainable practices across Europe. In addition to his environmental work, Szilárd has a deep understanding of foreign affairs, offering strategic advice and contributing to the development of policy initiatives in this field.
Cover photo: AI-generated illustration by Central European Affairs (CEA), 2026.
Support independent analysis and journalism at CEA Magazine: https://centraleuropeanaffairs.com/donation
For more insights on Central European political risk, EU institutional developments, and transatlantic relations, follow CEA Magazine and the CEA Talk podcast.
