In the spring of 2026, a study published by the CFE-CGC is shaking up the world of work: according to its projections, artificial intelligence could threaten up to 5 million jobs in France in the coming years. An impressive figure that raises as much concern as debate. But what does this number really hide? Should we panic, or seize the opportunities created by this technological revolution?

What does "5 million threatened jobs" really mean?

It is important to first nuance the word "threatened." The economists who use this expression are not talking about 5 million unemployed people tomorrow morning. They describe positions whose tasks could be partially or fully automated by generative AI over a 5 to 15-year horizon.

In other words, a job "exposed" to AI does not necessarily disappear: it transforms. An accountant will spend less time entering data and more time analyzing and advising. A lawyer will delegate documentary research to AI tools to focus on strategic counsel. The line between automation and augmentation is subtle, but it changes everything.

Which sectors are most exposed?

Unlike previous waves of automation (which mainly targeted workers), generative AI primarily affects highly skilled jobs with a strong cognitive component. The most concerned sectors in France are:

• Financial services: risk analysis, file processing, regulatory compliance
• Legal services: contract drafting, case law research, due diligence
• Human resources: resume screening, preliminary interviews, administrative management
• Marketing and communications: content writing, social media management, data analysis
• Customer service: handling standardized requests, first-level assistance

Conversely, professions that involve strong human interaction, non-reproducible creativity or a physical presence remain largely safe for now: craftspeople, healthcare workers, teachers, social workers, construction workers.

Is France well-prepared or not?

In terms of opportunities, France is holding its own. With more than 166,000 AI-related job offers published in 2026, it ranks among the top European countries. The government has also invested 2.5 billion euros in its national AI strategy, a significant portion of which is dedicated to training: the goal is to train 100,000 professionals by 2027.

The most in-demand jobs today reflect this transformation:

• Machine learning engineer
• Data scientist
• Prompt engineer
• AI ethicist
• Generative AI specialist
• Corporate AI governance manager

These are positions that did not exist five years ago and that today struggle to be filled due to a lack of trained candidates. The paradox of our time: while some fear losing their jobs because of AI, thousands of AI positions remain vacant.

What studies say about the real impact

The most serious studies on the subject agree on one point: AI will transform more jobs than it destroys. McKinsey estimated as early as 2023 that while 30% of tasks could be automated by 2030, less than 5% of professions would be entirely eliminated. In 2026, these projections are being confirmed in the field.

What is concretely happening, according to labor market observers, is a polarization: highly skilled and very low-skilled jobs hold up better than intermediate ones. A phenomenon already observed during previous waves of digitization, but accelerating with generative AI.

"AI has fairly quickly acquired the status of a technological revolution. But like all revolutions, it creates as much as it destroys — provided we prepare for it."

— Unédic Report 2026

How to prepare: practical advice

Faced with this transformation, what can you concretely do if you are an employee or in a career change? Here are some serious leads:

• Learn to work with AI, not against it. Mastering tools like ChatGPT, Copilot or Claude in your profession has become a differentiating skill.
• Develop what AI cannot do. Contextual judgment, empathy, singular creativity, ethical decision-making: these human qualities are gaining in value.
• Train continuously. The CPF (Personal Training Account) and Pro-A schemes allow everyone to fund AI training at all levels.
• Anticipate transformations in your sector. Talk with your employer about integrating AI into your processes — it is better to be an actor of change than to undergo a reorganization.

Conclusion: a threat to put in perspective, an opportunity to seize

The figure of 5 million "threatened" jobs is real, but it must be read with nuance. Economic history shows that every major technological revolution — from the steam engine to computing — has ultimately created more jobs than it destroyed, after a painful transition period for some.

The real question is not "Will AI take my job?" but "How can I evolve to stay relevant in a world augmented by AI?". In France, the tools exist: training programs, public investment, dynamic AI recruitment. It is up to each of us to take advantage of them.