AI amplifies the crisis of truth in business; beyond the tools, managers must train in critical judgment, methodical doubt and verification of sources.
With artificial intelligence (AI), we are entering a world where images, text, voice and even the apparent coherence of reasoning are no longer enough to objectify reality. Public debate focuses on productivity, technological sovereignty, global competition or the effects of automation on employment. But, for businesses, a risk remains underestimated. The blind spot is not just technical. It concerns the collective relationship to information, proof and the reality of facts, in particular with regard to science.
There is nothing impressionistic about this diagnosis. In 2024, the survey by the Jean-Jaurès Foundation, L’ObSoCo and Arte showed that 53% of French people suffered from information fatigue, including 38% “a lot”. Two years later, the phenomenon is not decreasing: 54% say they are tired of the information, including 39% “very tired”. Already in 2024, 77% said they sometimes limit or stop consulting the news, and 68% said they had deleted social media applications from their phone to regain control over their uses.
This evolution does not stop at the company’s doors. She goes in there every morning. The company was thinking of integrating skills; it must also deal with individuals immersed in an environment where the very credibility of the content has become problematic. And this is precisely why the question posed by AI is not only that of the performance of the tools. It is also that of the solidity of human judgment.
AI does not invent the crisis of truth, it amplifies it
Because AI is transforming the nature of intellectual work. It will no longer just be a matter of producing a note, a report or a summary, but of supervising content generated by AI agents. The value will no longer reside solely in execution, but in the ability to verify, correct, prioritize, contradict. In a word: judge.
This crisis of truth meets an already saturated world of work. The latest study by the Jean-Jaurès Foundation on information fatigue at work shows that one in four workers faces it, or nearly 7.5 million French people. The proportion rises to 39% among those who combine email, instant messaging and video. Even more telling: 62% of workers with a professional email box say they have already missed an important message buried among the others.
Training in AI is not enough: we must relearn how to judge
This is where the seesaw comes into play. The more systems are capable of producing plausible analyses, the greater the risk of confusing fluidity and reliability. An elegant mistake is still a mistake. A credible hallucination is no less a hallucination. To think that the technique will eventually self-correct is an illusion. The more powerful the tool becomes, the greater the need for human vigilance.
Therefore, the challenge for businesses is clear. Beyond AI training, we must learn not to give in to it and rehabilitate skills that we thought were secondary: reading carefully, cross-referencing sources, spotting bias, accepting complexity, asking for a rereading, etc.
As AI advances, the central issue is therefore not just equipment. It is also organized confidence, clarity of direction and the ability of leaders to explain. This is all the more true since the French no longer have confidence in traditional institutions, and the majority of them transfer this confidence to companies to develop their skills.
The company therefore has a decisive responsibility. It can no longer limit its ambition to technological equipment or increasing skills in tools. It must invest in education in information, in data, in methodical doubt. Not as a supplement of soul, but as a necessity of governance.
In the world that is opening up, the main competitive advantage will belong to those who know how to preserve this asset that has become rare: the ability to distinguish the probable from the false, the convincing from the true. The crisis of truth is no longer peripheral for businesses. It becomes one of their central questions.
