21SEP

Banque de France finds AI hiring gap

4 min read

16:45UTC

Banque de France economists found permanent hiring of under-30s in the most AI-exposed jobs 28% below its November 2022 level by March 2026, against 8% in the least exposed.

← Back to AI closes the hiring door, not the exit Jump to analysis ↓

EconomicAssessed

Key takeaway

Young people in France are hired far less often into the jobs AI can most readily do.

Economists at the Banque de France reported on 8 October that permanent-contract hiring of 21-to-29-year-olds in the most AI-exposed fifth of occupations was about 28% below its November 2022 level by March 2026, against 8% in the least exposed fifth⁠¹⁠². Exposure here means how readily generative AI tools can perform an occupation's tasks. At the start of 2026, hiring of young software developers and IT (information technology) engineers was 42% below November 2022, against 11% in less exposed occupations⁠³. The gap narrows for workers in their thirties and forties and nearly vanishes from 50.

The study, published on the Bank's blog under the authors' own names, counts declared CDI (contrat à durée indéterminée, France's open-ended employment contract) hires and terminations at about 6,000 firms in the Bank's business survey from January 2019, matched to an occupational exposure index for generative AI. November 2022, the base month, is when ChatGPT launched.

The Bank of Canada found the same shape in August, a job-finding gap that widened while dismissals barely moved. France finds it in declared hires at surveyed firms rather than a household survey. One difference: terminations of young exposed workers fell too, by 23% against 6%⁠⁴. Fewer under-30s in exposed jobs are hired, and fewer are let go, than in 2022; that matters because moving between jobs is usually how young workers win pay rises.

Danmarks Nationalbank reported in September that Danish firms adopting AI in 2023 had employment growth 11% lower than non-adopters two years on⁠⁵. A euro-area survey found AI use concentrated among graduates.

The French authors name other explanations. The European Central Bank (ECB) raised rates from July 2022 to September 2023, inside the same window, and the most exposed occupations are also the easiest to do from home, so post-pandemic remote hiring could mimic the pattern. Neither obviously explains why the gap fades with age.

Deep Analysis

In plain English

France's central bank, the Banque de France, studied how many young people get permanent jobs in different types of work. It compared jobs where AI tools can do much of the work, such as software development, with jobs where they cannot. Since late 2022, when AI chatbots became widely available, firms have hired far fewer under-30s on permanent contracts into the first group. Older workers in the same jobs are barely affected. The effect is on getting in, not on being pushed out.

Deep Analysis

Root Causes

France's employment protection makes a permanent contract a long bet. A CDI is costly to end outside a negotiated rupture conventionnelle, so a firm unsure which tasks generative AI will absorb has a reason to delay the hire rather than make it and reverse it later.

The study counts only CDI flows. Young French workers mostly enter through fixed-term contracts and, since the 2018 apprenticeship reform and the hiring subsidies that followed, through apprenticeships, so part of the missing CDI hiring may sit in those channels instead of vanishing.

The gap fading with age points to task content. Junior roles in coding, drafting and analysis are built from the routine tasks generative tools perform best, while senior roles in the same occupations carry client contact and sign-off that the tools do not replace.

What could happen next?

Consequence
Young French entrants to AI-exposed occupations may shift towards fixed-term contracts and apprenticeships while permanent hiring stays depressed.
Short term · Assessed
Risk
A smaller intake of junior software developers and IT engineers could leave French firms short of mid-career specialists by the early 2030s.
Long term · Suggested
Precedent
A third central bank, after Canada and Denmark, has placed the AI labour effect in hiring flows, which gives other statistics offices a template to test the same mechanism.
Medium term · Assessed

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: Japan's employment ice age, roughly 1993 to 2005, worked through the same door: large firms protected incumbent staff on lifetime contracts and cut graduate intake instead. Cohorts who missed a regular post at graduation stayed behind for decades, and the Japanese government launched a dedicated support programme for that generation in 2019.

Counter-parallel: When the Bell System mechanised telephone exchanges in the 1920s and 1930s, economists James Feigenbaum and Daniel Gross found it cut the hiring of new young operators while incumbents were largely spared. Later cohorts moved into secretarial and clerical work instead, so the entry-door loss did not leave a lasting scar.

The immediate cost falls on job-to-job mobility. With terminations of young exposed workers down 23% against 6% in less exposed jobs, fewer posts open up for the next cohort, and young workers lose the moves that usually deliver their largest pay rises.

Over five to ten years the pipeline of specialists could thin. If firms keep hiring experienced staff while cutting intake into software and engineering roles by around 42%, the supply of mid-career specialists in the early 2030s depends on training that no longer happens on the job. Lisa Kahn's work on American graduates who entered the 1980s recession found earnings losses that persisted for well over a decade.

Consensus view: The Banque de France authors, Palladino, Bergeaud, Bertheau, Bunel and Degos, join the Bank of Canada in finding the adjustment at the point of hire, not through dismissals. Danmarks Nationalbank found the same firm-level gap in September but said it saw no clear sign of an aggregate effect on Danish employment.

Counter-view: Oxford Economics argued in January that AI's role in job losses is overstated, because productivity has not accelerated the way large-scale labour replacement would require.

From Latin America, the Mexico, ¿cómo vamos? group found exposure concentrated in formal payroll work, so in economies with large informal sectors the same mechanism would reach far fewer young people.

Key tension: Whether the fall in terminations means firms are keeping experienced staff whose judgement now complements AI tools, or simply that nobody moves in a market cooled by the 2022-23 rate rises.

Sources:Banque de France·Blog du Modérateur·Central Banking

Mentions:France →ChatGPT →Bank of Canada →ECB →Mexico →Canada →Oxford Economics →Japan →Banque de France →Danmarks Nationalbank →Latin America →

First Reported In

Update #21 · AI marks French hiring and Accenture's 10-K

Banque de France· 22 Oct 2026

Read original →

Causes and effects

Caused by

Canada finds the hiring door narrowing

The Banque de France study found in French hiring data the exposed-versus-unexposed hiring gap the Bank of Canada found in August.

Occurred 24 Aug 2026

Read story →

Korea finds the missing youth jobs

The French result adds a second age-graded finding to the Bank of Korea's pension-record study of youth job losses in AI-exposed sectors.

Occurred 18 Aug 2026

Read story →

This Event

Banque de France finds AI hiring gap

It is the first euro-area evidence that AI exposure shows up in who gets hired, matching the Bank of Canada's August finding.

Different Perspectives

Korea's four largest banks, as reported by News1

KB Kookmin, Shinhan, Hana and Woori rehired 1,013 retirees last year, equal to 79.1% of their graduate intake, according to News1. Graduates face a smaller entry intake while returners take fixed-term credit-screening roles at lower pay, News1 reports.

High Court of Justice of Madrid

The court's employment chamber ruled in judgment 864/2026, dated 8 October and made public in early November, that introducing AI is not on its own a ground for dismissal. Employers must prove AI's effect through the four causes the Workers' Statute already lists.

Singapore Ministry of Manpower

The Singapore Ministry of Manpower put third-quarter retrenchments at 4,800, concentrated in information and communications, financial services and manufacturing, with business reorganisation the main driver. Its release does not mention AI.

Challenger, Gray & Christmas

Challenger reported on 5 November that market and economic conditions overtook AI as the leading stated reason for 2026's US job cuts, at 130,531 against 121,401. AI was cited for 1,265 cuts in October, its lowest month of the year.

Office for Budget Responsibility

The OBR's 28 October outlook showed output per worker in information and communication growing 7.6% a year, yet kept AI's boost to UK productivity at 0.2 percentage points a year. It does not attribute the faster growth to AI.

Central Bank of Ireland

Governor Gabriel Makhlouf said on 2 November that AI investment has supported Irish activity and flagged the debt financing behind it as a financial-stability risk. On jobs, he named skills bottlenecks and mismatches, not job losses.

Written with AI from Lowdown's published reporting. Not individually reviewed by an editor. Editorial standards.