From Intel (24,000) to UPS (20,000); Amazon (30,000) to Verizon (15,000); and even Bosch (13,000), the numbers keep piling up, fueling the idea of an economy where AI would gradually replace human labor. Some companies, like Salesforce and Amazon, no longer hesitate to explicitly attribute part of their workforce reductions to AI.

This phenomenon is global: it affects the United States, Europe, and even Asia, and resembles more a structural reshaping of the workforce than a cyclical adjustment. Initially constrained to administrative and support functions, AI is now extending its reach to more complex assignments.

Navigating structural transformation and financial trade-offs

Salesforce estimates that AI could handle up to 50% of its internal tasks, partly justifying its decision to implement layoffs. However, AI cannot be held solely responsible for all workforce reductions, as other factors come into play:

1. Post-pandemic over hiring: Certain sectors, such as consulting and technology, have engaged in large-scale recruitment. However, today’s economic slowdown calls for a realignment.
2. Pressure on margins: Companies are seeking to reduce costs and focus their resources on activities deemed most strategic.

It is therefore essential to distinguish cyclical job losses from those directly attributable to AI in order to accurately assess its impact on the labor market.

The Coué Method of AI: a new narrative facade

AI has now become a marketing buzzword, perceived as a guarantee of progress and performance. For some companies, however, this integration is more a smokescreen than a genuine strategic shift or growth driver.

According to the Global AI Confessions Report (March 2025) ^[1], 74% of CEOs fear for their jobs if they do not quickly demonstrate measurable business gains linked to AI. This pressure is fueling a surge in AI projects, even though surveyed executives estimate that around 35% of these initiatives are mere showcases: a phenomenon referred to as “AI Washing”.

Moreover, this AI race is accompanied by massive investments: this year, more than $380 billion^[2] is expected to be invested in AI infrastructure by tech giants. This raises a critical question: where will the resources come from to finance such colossal amounts?

At first glance, the soaring stock prices of tech companies might seem sufficient to attract the necessary funds. However, the dynamics of cross-shareholdings among major tech players, feeding each other^[3], create a bubble effect, forcing these companies to demonstrate the profitability of their AI projects quickly.

Yet, at this stage, actual productivity gains do not offset the costs incurred. This gap fuels a self-fulfilling logic: companies are compelled to symbolically validate the effectiveness of deployed technologies in order to maintain market confidence and prevent the bubble from bursting.

Thus, one might wonder whether these staff reductions have become one of the main levers to materialize the expected gains and reinforce the idea that AI is the source of these gains.

This logic goes beyond the tech sector: other companies are adopting the same narrative, claiming they are ‘laying off staff because AI has made them more efficient,’ even though the actual integration of these tools remains uncertain.

AI: What are the real benefits? The end of routine tasks? Jobs replaced or reinvented?

The gap between the promises surrounding AI and the actual value created is likely to lead to a gradual market correction. According to Gartner, up to 40%^[4] of agentic AI projects could be abandoned by 2027 due to excessive costs, uncertain benefits or poorly managed risks.

Nevertheless, the impact of AI on work organization is very real. It enables the automation of numerous repetitive and informational tasks (though not all), gradually transforming work practices. Yet employees often underestimate the speed at which these technologies advance.

A notable example: OpenAI has hired around one hundred former investment bankers to train an AI capable of performing complex financial modeling, a task historically assigned to junior analysts.

As a result, entry points into these careers are shrinking, even disappearing. Various studies show that companies heavily leveraging this technology now hire fewer young professionals than before. Between January 2024 and July 2025, job postings targeting young talent dropped by 24 percentage points^[5] globally.

Seniors and more experienced profiles remain less affected for now, but they could be subjected to a rapid shift in required skills.

Train rather than replace: A strategic imperative

Professionals must now: 1) Identify differentiating skills 2) Develop creativity, critical thinking, sector expertise and systems thinking 3) Keep track of developments in their industry and company 4) Learn to master AI tools

Mastering AI does not mean becoming a technical expert but rather being able to use it on a daily basis, just like traditional office tools. It also involves knowing how to challenge and verify the reliability and quality of data, avoiding biases in generated outputs, and adhering to confidentiality and professional ethics principles that govern its use.

Some companies are already tightening their requirements. For example, Accenture has laid off several thousand employees unable to integrate AI into their practices, while KPMG now requires every consultant to justify their daily use of AI.

AI is transforming the way we work: the challenge is to embrace these tools and invest in ongoing learning, not simply endure the shift.  The real risk is therefore not just job loss, but professional obsolescence.

Companies have a central role to play in this transition: they must ensure a just transition so that productivity gains do not result in deteriorating conditions for employees. This calls for constructive social dialogue, comprehensive training programs, safeguarding employability, ensuring fairness in exit plans and transparency in decision-making. Training should not be viewed as a cost, but as a strategic investment, a driver of stability and productivity essential for long-term success.

^[1] Muehmel, Kurt. « The High-Stakes Reality of Leading With AI: Confessions From 500 Global CEOs ». Dataiku Stories – Blog, 27 march 2025

^[2] Levy, Ari. « How much Google, Meta, Amazon and Microsoft are spending on AI – Tech’s $380 billion splurge: This quarter’s winners and losers of the AI spending boom ». CNBC, 31 oct. 2025

^[3] Représentation des valeurs de marché des entreprises du secteur de l’IA. Source: Bloomberg News reporting, 2025

^[4] Gartner. « Gartner Predicts Over 40 Percent of Agentic AI Projects Will Be Canceled by End of 2027 ». Gartner Newsroom – Press release, 25 june 2025

^[5] Randstad. « 24% Decline in Financial Services Junior Roles Spurs Shift to an AI-Driven, Skills-First Approach ». Randstad – Press release, 22 september 2025