Will AI Take My Job?

What the data actually shows

The honest picture is this: AI is real, it is affecting employment, the effects are uneven, and the trend is accelerating.

Current displacement. Goldman Sachs estimated in April 2026 that AI is responsible for the net displacement of approximately 16,000 U.S. jobs per month, or roughly 192,000 jobs annualized. This is the first time a major Wall Street research desk has attempted to isolate AI's contribution from broader macroeconomic factors. Goldman projects this monthly figure rising to 22,000-28,000 by the end of 2026 as multimodal AI agents move from pilot deployments to full implementation.

Long-term projections. Goldman's baseline estimate is that AI will eventually displace 6-7% of the U.S. workforce, approximately 11 million workers. Globally, their modeling suggests around 300 million full-time jobs could be affected by generative AI. The World Economic Forum's Future of Jobs Report projects net displacement of 14 million jobs by 2027. McKinsey estimates that 30-50% of current work activities could be automated depending on industry and pace of adoption.

The task vs job distinction matters. Most jobs are not being fully eliminated; most are being transformed. McKinsey research from late 2025 found that today's AI could theoretically automate about 57% of U.S. work hours, but full job elimination is much rarer than task automation. The pattern is that AI takes some parts of a job, leaves other parts to the human, and changes what the human's day looks like. Whether the changed job still requires the same number of people depends on whether productivity gains translate into reduced headcount.

The demographic concentration. Younger workers are hit hardest. Workers aged 22-25 in the most AI-exposed roles have seen a 6% employment decline from late 2022 to September 2025. Software developers aged 22-25 saw a nearly 20% decline from their late-2022 peak. Anthropic CEO Dario Amodei predicted in 2025 that AI could eliminate roughly 50% of white-collar entry-level positions within five years. Older workers in the same fields have remained largely stable; the pattern is generational, not occupational only.

The occupational concentration. Goldman Sachs identifies the most exposed occupations: computer programmers, accountants and auditors, legal and administrative assistants, and customer service representatives. The pattern is jobs heavy in routine cognitive work (writing, summarization, data entry, basic analysis, code completion). The pattern reverses the historical assumption that automation hits manual work first; this wave hits cognitive work first.

The geographic and gender concentration. One of the least-discussed findings: AI displacement risk is concentrated in administrative roles disproportionately held by young women, particularly in U.S. metropolitan areas where these roles are clustered. The Yale Budget Lab's September 2025 analysis and SHRM's 2025 Automation/AI Survey both found this pattern. The benefits of AI accrue largely to firms and shareholders; the costs concentrate on specific demographic groups in specific occupations in specific places.

The current state. The SHRM 2025 Automation/AI Survey, which sampled 20,262 U.S. workers, found that at least 50% of tasks are already automated in 15.1% of U.S. employment, representing approximately 23.2 million American jobs where automation has already crossed the majority threshold. A further 7.8% of employment (roughly 12 million jobs) involves workers using generative AI for more than half their tasks. This is not a future projection; this is now.

The honest summary: AI is affecting employment in real, measurable, accelerating ways. The effects are unevenly distributed by age, gender, occupation, and geography. The aggregate effect on total employment is still net negative but modest. The distributional effect on specific groups is severe.

The question secular analysis cannot answer

The data above tells us what AI is doing to jobs. It does not tell us what jobs are for. That is a different question, and most of the contemporary AI labor debate avoids it.

The implicit assumption in most policy discussions is that work is a means to an end: workers earn wages, which they use to obtain goods and services, which they consume to live the lives they want. On this view, if AI provides the goods and services without the workers, and society finds some way to distribute the gains (universal basic income, social safety nets, retraining programs), then we have solved the problem. The work itself does not matter; what matters is what work provides.

Catholic teaching has been arguing against this view for more than a century. The argument is that work is not merely a means; work is part of what human life is for. People are workers in a deep sense, not just because they need wages. Lose work, and you lose more than income. You lose a way persons participate in God's creative activity, develop their own capacities, support their families, and contribute to community. The economy exists for people, not people for the economy, but this does not mean people can simply be made happy with consumer goods while machines do the work.

This is the question Pope Leo XIV's Magnifica Humanitas is positioned to put back on the table. The encyclical was signed on the 135th anniversary of Rerum Novarum, the founding document of Catholic Social Teaching, which is fundamentally about labor. The signing date is a signal. The encyclical will likely insist that the AI labor transition cannot be evaluated only by its effects on aggregate productivity or even on aggregate employment. It must be evaluated by what it does to workers as persons, which is a different question.

What Catholic teaching has said about work

The Catholic tradition on work is rich and old. Three documents are most important for the AI conversation.

Rerum Novarum (Pope Leo XIII, 1891). The founding document of modern Catholic Social Teaching, written in response to the First Industrial Revolution. Leo XIII faced a world in which factory work had broken older patterns of craft, family, and community; workers were exploited; capital was concentrated; and two competing ideologies (laissez-faire capitalism and revolutionary socialism) offered competing solutions, both of which the Church found inadequate. Rerum Novarum articulated a Catholic third way: affirming workers' rights and dignity while defending private property, calling for just wages and unions, and grounding both arguments in the dignity of the human person. The encyclical's framework still defines how Catholic teaching engages economic questions.

Laborem Exercens (Pope John Paul II, 1981). The most thorough Catholic treatment of work as such. John Paul II distinguished between the objective dimension of work (what is produced; the output) and the subjective dimension of work (what work does to and for the worker as a person; the becoming). His central argument was that the subjective dimension takes priority. Work is for the worker before it is for the output. A society that maximizes output while degrading the workers who produce it has gotten its priorities backwards. The encyclical articulates a series of consequences: workers' rights to just wages, to organize, to participate in the direction of their work; the family as the primary social institution that work supports; the relationship between work and rest; the meaning of work as participation in God's creative activity.

Magnifica Humanitas (Pope Leo XIV, signed May 15, 2026; releases May 25, 2026). The first papal encyclical on AI. The signing date connects it directly to Rerum Novarum. The encyclical will almost certainly engage AI's effects on work substantially, building on both Rerum Novarum's social framework and Laborem Exercens's theology of work. The page on Magnifica Humanitas will be updated with the full text analysis after the May 25 release.

Beyond these three core documents, the broader tradition contributes substantially. Quadragesimo Anno (Pius XI, 1931) developed the principle of subsidiarity. Mater et Magistra (John XXIII, 1961) addressed economic development. Centesimus Annus (John Paul II, 1991) engaged the collapse of communism and articulated Catholic teaching's nuanced position on market economies. The Catholic Social Teaching primer covers the full tradition.

The convergent claim across these documents: work is central to human life, the economy exists for people rather than vice versa, and any technology or system that degrades workers as persons fails its most fundamental moral test, regardless of what efficiency gains it produces.

The subjective dimension applied to AI

John Paul II's distinction between the subjective and objective dimensions of work is the most useful single concept for evaluating AI's effects on labor. The question is not whether AI produces more output (it does) or whether AI is faster than humans at certain tasks (it is). The question is what AI is doing to workers as persons.

Several specific effects deserve attention.

Deskilling. When AI handles the cognitively challenging parts of work and leaves humans to handle routine inputs, outputs, or oversight, workers' skills can atrophy. The radiologist who reads with AI assistance gets better at the cases the AI flags, but loses practice on the cases it handles silently. The writer who drafts with AI loses some of the deep practice that made them a writer. The customer service representative who scripts through AI-generated suggestions loses the judgment that came from handling cases independently. Catholic teaching's concern is not that deskilling is inefficient (it may be efficient in the short run). The concern is that work is supposed to develop the worker, and work that does not develop the worker fails one of its primary purposes.

Surveillance and managerial control. AI enables monitoring of workers at granularities previously impossible. Keystroke logging, productivity scoring, conversational analysis, attention tracking. These tools turn workers into objects of optimization. Laborem Exercens's insistence on the priority of the subjective dimension is sharply violated when workers are treated as outputs to be tuned. The relevant moral test is not whether the surveillance increases productivity; it is whether it respects the workers as persons.

The disappearance of entry-level positions. The most acute current effect of AI on work is the elimination of the routine cognitive work that has historically constituted entry-level white-collar jobs. The 22-25 year old who would have started as a junior associate, a paralegal, a software developer, or a customer service representative now faces a job market in which the bottom rungs of the ladder have been removed. Older workers who completed those entry-level stages are stable; younger workers cannot get to the same place. This is what some observers have called "the apprenticeship crisis." Catholic teaching's concern is not just for the individual workers affected; it is for the long-term collapse of the formation pipeline that produces experienced workers, leaders, and craftspeople. The next decade's senior practitioners have to come from somewhere.

The concentration of gains. The productivity gains from AI accrue largely to firms and shareholders. The costs (displacement, retraining, disrupted careers) accrue largely to workers. This is not a new pattern in technological transitions, but it is sharper with AI than with most previous transitions because the gains are large and the affected workers' political power is weak. Catholic teaching has been making the argument that workers are owed a substantial share of productivity gains since Rerum Novarum. The argument has been disputed throughout, but it remains the Catholic position.

The loss of work itself. The most extreme case is workers who lose their jobs entirely and do not find equivalent ones. The proposed solutions to this scenario (universal basic income, retraining programs, expanded safety nets) address one dimension of the loss: the income dimension. They do not address the other dimensions Catholic teaching identifies: the loss of participation, of development, of contribution. A worker who receives income but has no work is materially supported but has lost something Catholic teaching insists is essential. This is not an argument against safety nets; it is an argument that the question of work cannot be solved by income alone.

What's likely in Magnifica Humanitas

The encyclical's full text releases May 25, 2026. Based on Pope Leo XIV's prior statements and the encyclical's framing, several themes are likely.

Explicit continuity with Rerum Novarum. The signing date and the choice of papal name both signal this. Expect the encyclical to reference Leo XIII's 1891 framework explicitly and to position itself as the next major social encyclical in that tradition.

The dignity of workers as the central moral concern. Pope Leo XIV's prior statements on AI consistently center the person, not the technology. The encyclical will likely insist that AI labor questions cannot be reduced to aggregate productivity, GDP, or even aggregate employment. The relevant moral test is what AI does to workers as persons.

Concrete economic concerns. Antiqua et Nova and Pope Leo XIV's prior addresses have addressed AI's concentration of economic power, the gig economy, algorithmic surveillance, and the disposability of workers in AI-enabled production. Expect these concerns to be developed at length.

The call for structures. Catholic teaching has consistently called for institutional structures (unions, regulations, just wages, social safety nets) that protect workers from market pressures. The encyclical will likely call for analogous structures for the AI age, possibly extending to questions about algorithmic transparency, worker participation in AI deployment decisions, and the distribution of AI productivity gains.

The spiritual dimension. Pope Leo XIV has emphasized that AI's effects on the interior life matter as much as its effects on the material conditions of work. Expect attention to how AI changes what work feels like, what workers become through their work, and whether the dignity of work as participation in God's creative activity can survive the AI transition.

A call to action. Encyclicals typically conclude with a pastoral call. The encyclical will likely call on Catholics, on employers, on workers, on policymakers, and on AI developers to take specific responsibility for shaping the transition. The expectation will not be that the Church can stop AI development; it will be that the development can be shaped, and that shaping it is a moral duty.

This page will be updated with the actual text analysis once Magnifica Humanitas releases.

What workers should actually do

Practical orientations based on current evidence.

Understand your specific exposure. The displacement risk varies enormously by occupation, age, geography, and the specific tasks your job comprises. Workers in high-exposure occupations (programming, accounting, legal and administrative support, customer service) need to take the transition more seriously than workers in low-exposure occupations (skilled trades, nursing, teaching, complex relational work). Goldman Sachs and Brookings have published occupation-level exposure analyses worth consulting for your specific role.

Learn to use AI tools well. The research is convergent on this: workers who use AI effectively tend to outperform both non-users and full-replacement automation. The pattern is that AI augments skilled workers and displaces unskilled ones. Becoming the kind of worker who uses AI well is one of the strongest defenses against being the kind of worker whom AI replaces.

Develop the capacities AI does not replicate. Relational work (caring for people, leading teams, navigating complex human situations). Complex judgment (cases where the right answer requires weighing values, not just optimizing functions). Physical work requiring presence (skilled trades, healthcare, infrastructure). Creative direction (deciding what is worth making, rather than executing the making). Workers who develop these capacities are not just defending against AI; they are positioning for the work that will become more valuable as AI handles other work.

Participate in collective conversations. Catholic teaching insists that workers' agency in shaping their own labor conditions is part of the dignity of work itself. This is not a question to leave to employers, AI labs, and policymakers alone. Engage with unions, professional associations, civic life, and democratic processes about how AI is deployed in your industry. Workers without organized voice tend to bear costs they did not consent to.

Take retraining seriously, but not naively. Retraining programs work for some workers in some transitions; they fail for many in many. The honest evidence is that retraining is one tool among several, not a complete solution. If you are in a high-exposure occupation, start the transition before you are forced to.

For young workers specifically. The entry-level white-collar pathway is genuinely disrupted. The traditional path (graduate, find a junior role, develop skills over five to ten years, advance into senior work) is harder to walk now than it was a decade ago. Strategies that worked for the previous generation may not work for yours. Consider apprenticeship-style learning, hands-on experience, and roles that develop capacities AI cannot replicate, even at the cost of starting salary.

What employers should actually do

If you are an employer, decision-maker, or board member thinking about AI deployment, several orientations follow from Catholic teaching.

Treat AI workforce planning as a senior strategic question, not an HR optimization. Goldman Sachs's April 2026 report observed that firms succeeding in the AI transition treat workforce planning as a C-suite priority. Firms that treat it as cost reduction make decisions they will regret in three to five years when they discover they have no pipeline of experienced workers.

Invest in retraining and redeployment before headcount reduction. The cost of replacing an experienced worker with an AI-augmented new hire is higher than most cost-benefit models acknowledge, because the experience itself has value the model does not capture. Catholic teaching's argument is that this is also a moral question, not just an economic one. Workers who have given years to a firm have legitimate claims on consideration as the firm's work changes.

Solve the apprenticeship crisis. Entry-level positions are not just costs to the firm; they are the pipeline through which experienced workers are produced. Eliminating them in favor of AI may look efficient in the short run; in five to ten years, the firm discovers it has no senior practitioners with the kind of grounded experience that AI cannot provide. Building pipelines requires intentional choices about retaining entry-level roles even when AI could handle the immediate tasks.

Treat surveillance and monitoring as moral questions. The AI-enabled granular tracking of workers is technically available; whether it is morally appropriate is a different question. Catholic teaching's emphasis on the dignity of workers gives a clear answer: monitoring that treats workers as outputs to be optimized fails. Monitoring that supports workers' development and wellbeing is a different category. The distinction often matters more than the technical capability.

Distribute productivity gains. AI is producing real productivity gains. The question of who captures those gains is a moral question, not a technical one. Catholic teaching has consistently held that workers are owed a substantial share of productivity gains. Firms that retain all the gains may be technically permitted to do so by current labor markets; they are not behaving in line with the Catholic tradition's view of what employers owe their workers.

What's really at stake

The AI labor transition is not just an economic question. It is a question about what kind of society we are building.

The technical narrative says AI will increase productivity, and the welfare state can manage the transitional costs. This is partly true, and partly an evasion. It treats work as a problem to be solved rather than a good to be sustained. It assumes that displaced workers can be made whole through income transfers, training programs, or new sectors. Sometimes they can. Often they cannot. The evidence from previous technological transitions (the deindustrialization of the American Midwest, the collapse of various skilled trades, the decline of manufacturing employment) is that affected workers and their communities often suffer for decades. The AI transition has no guarantee of being better.

Catholic teaching's argument is that the AI transition cannot be managed as a purely technical or economic question because work is not a purely technical or economic thing. Work is part of what human life is for. The transition has to be evaluated by what it does to workers as persons, to families that depend on workers' wages and presence, to communities that are organized around shared labor, and to the broader culture that is formed by what work means.

This is the conversation Magnifica Humanitas is positioned to put back into mainstream discussion. Pope Leo XIV's choice of papal name, his signing of the encyclical on the Rerum Novarum anniversary, and his consistent framing of AI as the new industrial transformation all point to the same claim: the AI labor transition is the central moral question of our economic moment, and the Church has resources from more than a century of social teaching to bring to it.

Workers and employers reading this page can engage that conversation now, before the encyclical releases. The frameworks are old; the questions are new. The honest answer to "will AI take my job" is: maybe, partly, eventually, depending. The harder answer is: regardless of what happens to specific jobs, the question of what work is for, and how to honor the dignity of workers in a changing economy, will not be solved by AI or its absence. It will be solved, if at all, by sustained moral attention from the people building, deploying, regulating, and performing the work of the coming decade.

You are one of those people. Catholic teaching has something useful to say about the moral seriousness of the decisions in front of you.