Laboratory of Innovation and Future Transformations With the support of Accenture and Google
  • Home
  • Publications
  • AI trackers
    • Overview

    • AI diffusion (jobs)
    • AI diffusion (science)
  • Innovation data
    • Overview

    • Productivity
    • High-tech innovation
    • Breakthrough innovation
  • Members
  • FR|EN

AI diffusion (jobs)

Real-time monitoring of AI adoption in job postings in France

AI
Jobs
Interactive dashboard analyzing the diffusion of artificial intelligence through French job postings, based on the JOCAS database (DARES).

Total AI job postings

...

...

AI firms

...

...

Generative AI job postings

...

...

GenAI firms

...

...

Trend in AI-related job postings

Total

AI adoption by sector

Total
Generative AI

Sector share of published job postings

Distribution of adopters by firm size

AI

AI job postings by occupation (ROME)

AI

Trend in skills in AI job postings

Map of French employment zones

Total
AI

Top 10 employment zones

Click on a zone to see the details

Working conditions (all job postings)

Share (%)
Source: Data extracted from the JOCAS database (DARES) and analyzed with machine learning models to detect and classify AI-related job postings.

About this tracker

This dashboard tracks the diffusion of artificial intelligence in the French economy in real time through the analysis of job postings.

Indicators tracked

  • Total AI job postings : cumulative number of job postings mentioning AI skills
  • Generative vs. traditional AI : distinction between job postings targeting generative AI (GPT, LLMs, etc.) and traditional AI (ML, vision, NLP)
  • Sector distribution : breakdown between industry and services
  • Occupations (ROME) : classification of job postings by occupational family (ROME, the French repertoire of occupations)
  • Technical skills : trend in the technologies in demand (TensorFlow, PyTorch, etc.)
  • Geography : breakdown by employment zone

Methodology

The analysis relies on:

  1. Collection : continuous extraction of job postings from the JOCAS database (DARES)
  2. ML classification : machine learning models trained to detect mentions of AI
  3. Categorization : automatic distinction between generative AI, computer vision, NLP, etc.
  4. Aggregation : computation of monthly and quarterly indicators

Authors

This tracker is developed by:

  • Paul Delbouve — Collège de France, PSE, DGE
  • Antonin Bergeaud — HEC Paris, LSE, CEPR

Related publication

The diffusion of artificial intelligence in France Aghion, P., Bergeaud, A., Bunel, S., & Delbouve, P. (2025) Économie et Statistique/Economics and Statistics (forthcoming)

NoteCitation

If you use the data from this tracker, please cite:

@techreport{aghion2025diffusion,
  title     = {Diffusion de l'Intelligence Artificielle en France},
  author    = {Aghion, Philippe and Bergeaud, Antonin and Bunel, Simon and Delbouve, Paul},
  year      = {2025},
  institution = {Laboratoire de l'Innovation et du Futur du Travail},
  type      = {Working Paper},
  number    = {2025-1}
}

Update

The data are updated quarterly.

Laboratory of Innovation and Future Transformations HEC Paris · Hi! Paris

  • Publications

  • AI trackers

  • Data

  • Team

  • About

© 2026 Laboratory of Innovation

HEC Paris Hi! Paris Île-de-France Region IEI Google Accenture