Artificial intelligence chips away at jobs, France absorbs the shock

Introduction: a report nobody really wanted to read

A number that hits you square in the face

Last month, the Direction générale du Trésor (France's Treasury Department) published a note that should have made the front page of every economic newspaper in the country. Titled "Artificial intelligence, what effects on employment?", this study, signed by economists Martin Chopard, Elisa Cotet, Tristan Gantois, and Eloïse Villani, delivers a finding the public has already been feeling in its bones for months, according to the Direction générale du Trésor.

According to the document, 62% of French people already view artificial intelligence as a direct threat to their job, a figure drawn from the 2025 digital barometer. This isn't a vague hunch: it's a clear majority of the population living with this worry daily, between two meetings and a piece of software that seems to do a little more of their job every month.

A technical note hiding very real anxiety

Behind the muted language of Treasury economists lies a much rougher reality. The report acknowledges that the first waves of layoffs attributed to artificial intelligence are already concentrated in the most exposed sectors, notably finance and IT, even though AI sometimes serves as a pretext to mask other reasons for staff cuts.

That may be the most honest sentence in the entire document: no one, not even the state, yet knows for certain where this wave will stop. And yet, this official uncertainty doesn't reassure anyone among those who dread, every morning, watching their position disappear.

Young people, the silent first victims of the transition

A slowing entry into the workforce

The Treasury report flags a particularly worrying phenomenon: young people's entry into the job market appears to be slowing since the rise of generative AI. This finding echoes observations from INSEE (France's national statistics agency), which documented a decline in employment among 22-to-25-year-olds in the occupations most exposed to automation, on the order of 16%, according to a note published in March 2026.

In concrete terms, this means young graduates looking for their first job in IT or programming are running into a job market that is hiring less than before, not because economic activity is slowing, but because entry-level tasks are increasingly automated.

The IT sector, a real-world laboratory

Between the fourth quarter of 2023 and the fourth quarter of 2025, employment in the information and communication sector fell by 3.0% in France, according to INSEE data. This decline comes even as the sector's value added keeps rising, a paradox that neatly illustrates the mechanism at work: producing more with fewer hands.

This pattern isn't unique to France: the United States shows a comparable dynamic, where digital investment drives growth without employment rising at the same pace, according to the same INSEE note.

Between 5% and 60% of jobs exposed: an admission of massive uncertainty

A range that says a lot about collective ignorance

The Treasury cites available estimates placing between 5% and 60% the share of jobs potentially exposed, in the short or medium term, to generative artificial intelligence. A range that wide isn't a technicality: it's an admission that no one today truly grasps the scale of the upheaval underway.

For France specifically, a study by Arquié et al. (2026), cited by the Treasury, estimates that 3.8% of work content is currently at risk of automation by generative AI, a figure that could climb to 16.3% within two to five years.

Five million jobs in the crosshairs

That same study, conducted jointly by Coface and the Observatoire des Emplois Menacés, estimated that around five million positions out of a workforce of thirty million people could be affected by this growing exposure to generative AI. Notably: unlike previous waves of automation, it's high salaries that appear most exposed, with an exposure rate of 22.1% for the top 10% of earners.

This reversal upends familiar assumptions: for decades, skilled workers were promised their degree would protect them from automation. That clearly no longer holds, and this reality deserves to be stated plainly to the young people investing today in long years of study.

PwC's global barometer confirms a two-speed labor market

Productivity soaring for a minority of companies

PwC's Global AI Jobs Barometer 2026, published in June, analyzes more than a billion job postings across six continents to reach a finding similar to the French Treasury's: AI is creating a two-speed labor market, where judgment and leadership are becoming even more highly valued skills, according to PwC.

According to that same barometer, the top 20% of companies most exposed to AI recorded labor productivity growth of 163% compared to a 2018 baseline, a considerable gap versus the rest of the global economy.

Job creation figures that mask targeted job losses

Other data, notably compiled by specialized platforms like Snowflake, Anthropic, or LinkedIn, show that 77% of organizations worldwide report net job creation linked to AI, often accompanied by targeted losses in specific technical and administrative roles. France fits this global trend, with a gradual reallocation of talent rather than a sharp, net destruction of jobs.

This nuance matters, but it shouldn't be used to downplay the very real pain of those whose position concretely disappears, even if another position, elsewhere, is created in parallel at a department or company they may never join.

France facing a choice: invest or fall behind

International competitiveness as the central argument

The Treasury is clear on one point: amid international competition, investing in artificial intelligence is becoming necessary to sustain competitiveness, and therefore employment, which will increasingly hinge on adopting this technology. In other words, France cannot afford to fall back simply because the transition is frightening.

This logic fits within the national France 2030 strategy, which sets ambitious targets: 16% growth over ten years driven by AI and a 20% productivity gain, according to data reported in several recent economic analyses.

Massive hiring needs despite it all

Paradoxically, a survey of French companies' labor needs, published on April 21, 2026, counts 2.2 million planned hires this year, including 84,000 in the tech sector alone, where 49% of companies nonetheless struggle to fill positions. The healthcare sector, meanwhile, plans 322,000 hires, largely spared from the direct threat of automation.

This contrast between labor shortages in some occupations and fear of replacement in others illustrates the real complexity of this transition, far removed from simplistic narratives of robots purely and simply replacing humans.

The European Central Bank also sounds the alarm

Five million jobs cited by 2030

The European Central Bank acknowledges the positive effects of artificial intelligence on the French economy while warning of negative short-term effects: up to five million jobs could be eliminated in France by 2030 due to AI-driven automation, a figure that aligns with projections from Coface and the Observatoire des Emplois Menacés.

The ECB also highlights a worrying structural risk: job losses could temporarily outpace job creation, with a cascading effect on tax revenue, since the most threatened positions are often the most skilled and highest paid.

Unemployment already rising, without a systematic direct link

French unemployment rose 1.7% year-on-year in 2026, without this increase being fully and directly attributable to artificial intelligence, according to data reported by France Travail (the national employment agency). This nuance matters: it's a reminder that AI is not the sole variable in a much broader economic equation, also shaped by international competition and multiple sectoral shifts.

But this nuance shouldn't serve as an excuse for inaction: even a partial contribution from AI to rising unemployment deserves a structured public response, before the phenomenon worsens further.

Training, the only credible response to the uncertainty

What economic history teaches about technological revolutions

The Treasury recalls a lesson from previous technological revolutions: the jobs created by new technology have, in the past, ended up offsetting those that disappeared, while the very content of occupations was deeply transformed. Nothing guarantees this pattern will repeat identically with generative artificial intelligence, but history at least offers a cautious frame of reference.

This time, however, the pace of change appears markedly faster than that of previous industrial revolutions, leaving less time for workers and institutions to adapt gradually, a factor several economists consider decisive going forward.

Public support deemed essential by experts

The Treasury report is explicit: support through public policy, particularly on training, is deemed necessary to help retrain workers whose jobs are threatened, and to encourage the spread of AI across the French economy as a whole. This isn't wishful thinking: it's a condition laid out in black and white by the state's own economists.

It remains to be seen whether that support will actually follow, concretely, on the ground, in vocational training budgets and retraining programs, or whether it will remain a good intention recorded in yet another report, quickly forgotten after publication.

Manual and healthcare jobs, a relative but not eternal refuge

Sectors spared by the wave, for now

Some sectors are weathering the push of generative AI automation better than others. Healthcare, with 322,000 planned hires in 2026, hospitality and food service, and construction remain largely spared, according to data reported by several recent studies on the French labor market. Manual occupations, which require physical presence and constant adaptation to unpredictable situations, remain for now out of reach of large language models.

This relative protection should not, however, be mistaken for permanent immunity: rapid advances in robotics combined with AI could, in the medium term, extend automation to tasks currently considered out of reach.

Human skill as the last line of defense

Experts consulted across several studies agree on one point: occupations requiring judgment, empathy, and the ability to adapt unpredictably remain, for now, the best protected against automation, regardless of the pay level attached to those positions. That's relatively good news for nurses, educators, and skilled tradespeople.

But this relative protection alone isn't enough to solve the overall equation of French employment, since not every worker threatened by automation can simply retrain overnight into one of these protected sectors.

Conclusion: a transition that won't forgive standing still

A report that raises more questions than it answers

The Treasury's note on artificial intelligence and employment offers no definitive prediction, and that's precisely what makes it credible. Between 5% and 60% of jobs exposed, between job creation and destruction that may or may not offset each other depending on the scenario chosen, the only certainty that emerges is that of a profound transformation of the French labor market already underway.

What young graduates struggling to land their first job are living through today, or what experienced employees seeing their position redefined or eliminated are going through, won't wait for statisticians to finally settle the debate between competing economic studies.

Political time versus technological time

Between the pace of government reports and that of technology adoption within companies, a worrying gap is opening up. France still has a window to invest massively in training and support, but that window is closing as generative artificial intelligence keeps advancing the capabilities it offers companies.

By Maxime Marquette, columnist

Columnist's transparency note

Who I am and my acknowledged biases

I sign this testimony under the name Maxime Marquette. I closely follow technological change, and I observe, in my own daily work, the concrete effects of artificial intelligence on content-production occupations, which inevitably shapes my perspective on this topic.

I am not an economist, and I rely exclusively on recognized public reports, notably those of the Direction générale du Trésor, INSEE, PwC, and the European Central Bank, systematically flagging discrepancies between their respective estimates.

What I don't know and my method

I don't know with certainty what exact share of French jobs will ultimately be affected by artificial intelligence, as estimates vary considerably from one study to another, ranging from 5% to 60% depending on the methodology used. My method consists of always attributing every number to its precise source, never merging divergent estimates into a single invented figure.

Sources

Primary sources

Direction générale du Trésor — Artificial intelligence, what effects on employment, note no. 391, June 2026

PwC — Global AI Jobs Barometer 2026

INSEE — With the rise of generative artificial intelligence, digital investment is rising, March 2026

Secondary sources

Anthem Création — AI and jobs in France: 5 million at risk, says Coface/OEM, March 21, 2026

Expat.com — How AI is reshaping the French job market, May 5, 2026

Adapte-toi — McKinsey 2026 report: AI's impact on French employment, April 23, 2026