AI · WORK · PREVENTION

Understand and prevent the effects of AI on occupational health.

Independent guidance for understanding how AI systems transform work, and how to anticipate their effects on health.

01 — THE PROBLEM

How does AI affect
occupational health?

Artificial intelligence can change tasks, deadlines, room for judgement, responsibilities and skills, with possible effects on occupational health.

These effects depend not only on technical performance, but on how a system is deployed, controlled and discussed. Prevention therefore means examining real work before deployment and after use begins: cognitive load, autonomy, surveillance, work quality, collective work and health. This concerns employers, worker representatives, public authorities and occupational safety and health professionals together.

FOR YOU

See how this question concerns you.

Assess a project
The Consummation of Empire, a Thomas Cole painting showing a city at the height of its power

A WARNING

Thomas Cole
The Consummation of Empire, 1836

In Cole’s series, this city at the height of its power immediately precedes chaos. It reminds us how quickly an equilibrium that seems secure can become fragile and tip if we stop paying attention. View the work ↗

02 — THE ARGUMENT

A model never acts alone.

The health effects of an AI system do not come from the model alone. They emerge from the way its capabilities are integrated into the organisation of work.

  1. 01

    Model

    • Capabilities
    • Errors
    • Data
  2. 02

    Work system

    • Targets
    • Deadlines
    • Control · Staffing
  3. 03

    Real activity

    • Workload
    • Autonomy
    • Skills
  4. 04

    Health

    • Physical risks
    • Psychosocial risks

04 — RESEARCH

Recent work

SCIENTIFIC ARTICLE / 2025

Large language models and emerging psychosocial issues at work.

Charles Broutin

The article examines the introduction of LLMs through the six Gollac factors: intensity, emotional demands, autonomy, social relationships, value conflicts and insecurity.

Open the DOI ↗

STUDY / JUNE 2026

Artificial intelligence to improve the precision of occupational physician recommendations.

Charles Broutin

An evaluation of a language model used as a second reader to identify wording defects, without replacing work-post analysis or medical judgement.

Read on the INRS website ↗

METHOD / 2026

Assessing the psychosocial effects of an AI project before and after deployment.

Charles Broutin

An exploratory method for describing workload, autonomy, skills, collective work and the conditions for a reversible trial separately.

Read the method →
View all research →
Professional portrait of Dr Charles Broutin
Charles Broutin, occupational physician

05 — ABOUT

An occupational physician’s perspective on AI deployment.

AI & Occupational Health is an independent editorial initiative led by Dr Charles Broutin, an occupational physician and AI lead for the French Society of Occupational Health. The project connects the evaluation of AI systems with the analysis of real work: what tools enable, what they shift within an organisation and what must be verified to protect health.

Learn more →

MONITORING

Follow how AI is transforming work.

Independent monitoring of uses, research and decisions that reshape occupations, responsibilities and prevention.

Subscribe to the newsletter ↗