IA washing: How companies justify layoffs with artificial intelligence

At the start of 2026, three announcements shook the French tech sector within a few weeks. Capgemini is cutting 2,409 jobs in France, equivalent to 7% of its French workforce. IBM France is slashing 300 positions, or around 10% of its local headcount. Accenture, already in the midst of a worldwide restructuring plan cutting 19,000 jobs since 2023, continues its overhaul and consistently cites digital transformation and AI as key reasons. In all three cases, the term “artificial intelligence” features prominently in official communications.

What deserves our attention is less the fact that jobs are being cut, and more the way these job losses are being justified.

When an IT services company lays people off, it’s rarely a surprise. But when it presents AI as the main cause, that’s when deeper analysis is required.

What Is Really Happening in France in Early 2026

Let’s look at the facts. Capgemini announced in January 2026 a “job protection plan” involving 2,409 positions in France. Management explicitly cited “the rise of artificial intelligence” and the need to “refocus skills.”

On paper, this is the perfect justification for a group that sells digital transformation consulting services to its clients.

Except Capgemini has also published its financial results. The company has faced squeezed margins since 2023, negative organic growth in the third quarter of 2024, and increased competition from Indian offshore service providers, namely Infosys, TCS, and Wipro.

These three players offer equivalent services at costs 40–60% below French rates. AI gets mentioned. Offshore is conveniently left out.

IBM France follows a similar path, with a telling history. This 300-job cut in 2026 is part of a series of restructurings going back to 2021, 2022, and 2023. Each time, the explanation evolves: first cloud, then cybersecurity, now generative AI.

Meanwhile, IBM’s French subsidiary declared 87 million euros in net profit in 2024. Layoffs at a profitable company that invokes every new tech trend as justification: it’s a noteworthy pattern.

The case of Onclusive stands apart for its frankness. This media analytics company executed, in 2025, France’s first official “job protection plan” (PSE) explicitly motivated by AI deployment. No rhetorical fluff: automated analysis tools directly replaced human analysts.

It’s harsh, but honest. Paradoxically, this concrete case better illustrates the real-world impact of AI than Capgemini’s or IBM’s grand statements.

Meanwhile, ASML, the Dutch lithography company essential to the global semiconductor industry, announced 2,000 job cuts at the end of 2024 without mentioning AI even once.

The reason given: a cyclical adjustment in demand, mostly in China. When a company as technologically advanced as ASML restructures without waving the AI flag, it puts the messaging from French IT services firms into perspective.

Takeaway: Three major tech companies cite AI to justify their social plans in France in early 2026. In at least two out of three cases, financial data and offshore competition provide more convincing structural explanations than disruption by AI.

What Is IA Washing?

IA washing is when an organization invokes artificial intelligence as the cause, driver, or justification for a decision, product, or strategy, while in reality, AI only plays a minor, secondary, or nonexistent role in that operational decision.

Applied to restructuring, it means presenting job cuts prompted by classic economic reasons (margin pressure, offshore optimization, slowing business activity) as the inevitable consequence of the AI revolution.

The phenomenon goes far beyond layoffs. Startups raise funds by “adding AI” to their pitch—even if their product is little more than an OpenAI API. Large companies rebrand existing tools with the “AI-powered” label.

HR managers talk about “AI augmentation” to describe layoffs paired with a two-day ChatGPT training session.

Why do companies do it? There are three main reasons.

• The first is legal: under French law, a job protection plan is governed by Article L.1233-3 of the Labor Code, which requires a “genuine and serious” economic reason. Citing a major technological shift makes the economic rationale more robust legally.
• The second is financial: markets reward companies that position themselves as players in AI. Announcing a restructuring plan “to accelerate AI transformation” is viewed more positively by investors than admitting a loss of competitiveness to offshore rivals.
• The third is narrative: for employees and the public, “AI is forcing us to adapt” is a less stigmatizing story than “we’re losing market share to cheaper service providers.”

How to Tell Real AI Impact from Empty Rhetoric

Here’s a five-point checklist to assess the sincerity of an AI justification in a restructuring plan.

Criterion 1: Was AI Deployed Before the Layoffs?

A genuine AI impact means tools have been rolled out, adopted, and have actually changed work processes before jobs are cut.

If a company announces AI deployment and job cuts simultaneously, the cause-and-effect sequence is questionable.

At Onclusive, automation tools were up and running for several months before the layoffs. At Capgemini, the “rise of AI” remains a strategic goal presented as if it’s already been achieved.

Criterion 2: Do Eliminated Positions Match Automatable Tasks?

AI mainly automates repetitive, codifiable, high-volume tasks. If job cuts hit junior production workers, basic data analysts, or routine software developers, the correlation is credible.

If cuts indiscriminately target managers, sales, or senior experts with no link to tasks likely to be automated, the AI argument falls flat.

Criterion 3: Is Offshore Mentioned or Omitted?

In IT services, competition from India, Morocco, or Romania is often the main driver of restructuring in France.

If a company cuts jobs in France while expanding its centers in India and leaves that out of their HR message, the omission is telling.

Capgemini employs more than 200,000 people in India. This context should be acknowledged.

Criterion 4: Do Financial Results Support the Stated Reason?

A profitable company cutting jobs “because of AI” deserves more skepticism than a company with lasting losses forced to adapt.

IBM France posted €87 million in net profit while restructuring: the need for urgent economic action is much less obvious than the narrative suggests.

Criterion 5: Is the Company Hiring for the Same Roles It’s Cutting?

The case of Amazon is revealing. Andy Jassy announced job cuts in 2024, citing automation through AI, only for Amazon to post thousands of job openings for nearly identical roles soon after.

This sort of contradiction often points to a geographic or salary-based restructuring cloaked as technological transformation.

What the Data Says About Employment and AI

The available data at the beginning of 2026 paints a more nuanced picture than corporate messaging. Challenger, Gray & Christmas tracked over 55,000 layoffs explicitly attributed to AI in the United States in 2024 alone, from among hundreds of thousands overall. The proportion remains small, but is growing.

The Forrester report from January 2026 identifies three types of AI impact on jobs:

direct eliminations (fully automated tasks), changed job roles (positions remain but change substantially), and net new jobs (roles created by AI).

According to Forrester, job transformations are by far the most frequent category, ahead of outright job losses. That doesn’t mean losses are negligible, however.

The PwC AI Jobs Barometer measures “AI penetration” by sector and cross-references this data with employment trends.

The result: sectors with high AI adoption see faster wage growth—but also a bigger wage gap between highly skilled and less skilled workers.

AI creates value for those who can use it; it erodes the position of those performing directly competing tasks.

The OECD estimates that 27% of jobs in member countries are highly exposed to automation by AI.

In France, the most exposed sectors are administrative services, accounting, customer support, and some standardized legal functions.

But “exposed” does not mean “to be eliminated tomorrow”: exposure measures the technical feasibility of automation, not its real-world deployment speed.

Sam Altman, CEO of OpenAI, stated publicly in 2024 that “AI will eliminate most jobs as we know them today.”

This quote is widely circulated. Less shared: in the same interview, Altman added that the timeline remains uncertain and new types of work will emerge.

The chopped-up quote fuels doomsday headlines; the full one calls for a more balanced analysis.

The analogy with Wassily Leontief is worth noting. In 1982, this Nobel-winning economist predicted that machines would make human workers obsolete, just as horses were replaced by combustion engines.

Forty years later, employment rates in developed countries have generally increased—despite (or thanks to) automation.

This isn’t an argument for downplaying the impact of today’s AI, which is unprecedented in its ability to compete with cognitive work.

It’s a reminder that predictions of mass job destruction have regularly overestimated the speed of change and underestimated the labor market’s ability to adapt.

What This Concretely Means if You’re in France

For employees or jobseekers in France, two practical points require your attention.

On the legal side, if you’re subject to a PSE citing AI, Article L.1233-3 of the Labor Code requires the economic reason to be genuine and serious.

Technological transformation can be a valid reason—but it must be documented: which tools were deployed, in which roles, with what measurable effect on workloads?

A lawyer specializing in labor law can demand your employer provide specific justifications. “AI will transform our sector” is too vague by itself to meet legal standards.

As for the European AI Act, whose most stringent provisions take effect in August 2026, AI systems used in HR decisions (hiring, evaluation, terminations) are classified as high risk.

Companies will need to document automated decisions and guarantee genuine human oversight.

This regulatory framework doesn’t ban layoffs, but it does impose greater traceability on AI-based decisions—making pure rhetorical justification harder.

However, describing the AI Act as a sufficient safeguard would be an overstatement: the regulation covers AI systems in HR—not the broader economic decisions behind restructuring, which remain under national labor law.

For workers wondering about the real impact of AI on their jobs, the distinction between technical exposure and actual deployment is key. A role may be technically automatable without the organization having the resources, will, or timeline to automate within the next two years.

And companies that actually deploy AI in production often redesign jobs before eliminating them, since the transition itself creates work.

The regulation of AI in France now revolves around the AI Act and the growing role of the CNIL (French Data Protection Authority) in overseeing automated decision systems in HR.

Social and Economic Committees (CSEs) have the right to be informed and consulted before any deployment of AI tools that affects working conditions—a right that’s often underused.

IA Washing: A Risk to Trust

The issue with IA washing isn’t just the inaccuracy of company messaging. It creates two equally harmful side effects.

First effect: when companies artificially inflate the role of AI in their restructurings, it fuels exaggerated anxiety among employees.

Jobs not actually threatened by AI are perceived as being at risk. Costly, rushed career changes may result from a false sense of urgency. The fear generated exceeds the real impact.

Second, and just as bad: when employees realize AI was only a rhetorical pretext, mistrust takes root about future, real technological shifts.

If AI has been used to justify typical cost-cutting, why believe the next warnings about real automation? IA washing undermines our collective ability to prepare for genuine transformations.

Data from Forrester, the OECD, and PwC all point to the same thing: AI will indeed significantly change professions over the next five to ten years, though unevenly by sector. This real transformation deserves honest debate, targeted training policies, and an appropriate legal framework.

It does not deserve to be used as a fig leaf for economic decisions rooted elsewhere.

Distinguishing signal from noise, here as elsewhere, remains the most crucial skill.

For employees, union representatives, journalists, and anyone making career decisions in the coming years based on what companies say about AI.

Key point: IA washing in restructuring plans causes two kinds of harm: excessive, short-term anxiety among workers not actually at risk and lasting distrust that makes it harder to prepare for authentic technological change.

If you want to keep up with the evolution of the AI Act and the concrete implications for jobs and HR in France, subscribe to our newsletter.

FAQ

What exactly is IA washing, in one sentence?

IA washing is invoking artificial intelligence as the main reason or justification for a corporate decision, when in reality AI plays only a secondary or non-existent role in that decision. In the context of restructuring plans, it means cloaking job cuts—driven by classic economic reasons—in technological language.

Is Capgemini’s restructuring plan really caused by AI?

The honest answer: partially, perhaps. Capgemini is facing real competitive pressure from Indian offshore providers, negative organic growth since 2023, and squeezed profit margins. AI is indeed changing productivity in some software development areas. But presenting AI as the sole or main cause is hard to prove with the public information available.

What does Article L.1233-3 of the Labor Code say about layoffs for technological reasons?

This article defines economic layoffs and includes technological change as a valid reason, alongside economic difficulties and necessary reorganizations for competitiveness. The cited technological change must be real, well documented, and directly impact the jobs being cut. A vague statement about “AI-driven sector transformation” with no concrete, documented deployment can be challenged.

Does the European AI Act protect employees from AI-related layoffs?

The AI Act categorizes AI systems used in HR decisions (hiring, evaluation, firing) as high-risk, requiring transparency, human oversight, and documentation. However, the AI Act regulates the use of AI in HR processes—not the economic restructuring decisions themselves, which are still governed by national labor law. The two legal frameworks are complementary, but distinct.

What’s the difference between an AI “exposed” job and a job eliminated by AI?

Exposure refers to the technical feasibility of automation: could the tasks of this position theoretically be handled by an AI system? Actual elimination depends on real-world deployment, transition costs, management decisions, and timing. The OECD estimates 27% of jobs are highly exposed in France, but technical exposure frequently precedes actual job loss by several years.

Why do companies prefer to cite AI rather than offshoring to justify layoffs?

The offshore argument is seen as an admission of lost competitiveness and triggers a more negative reaction from the media and unions. AI, by contrast, is presented as an inevitable external factor—less stigmatizing for management and more palatable for financial markets. Legally, both reasons may validate a restructuring plan, but the AI narrative is more attractive in 2025–2026.

What does the January 2026 Forrester report say about jobs and AI?

Forrester distinguishes three impact categories: direct elimination of jobs by AI, job transformations where the role changes significantly but is not lost, and net new positions. Job transformations are by far the largest group numerically, arguing in favor of reskilling policies rather than just severance plans for laid-off workers.

How does the Onclusive case differ from Capgemini or IBM?

Onclusive rolled out France’s first explicitly documented restructuring plan motivated by AI deployment: automated media content analysis tools were implemented and field-tested before the layoffs, and the analysts cut were performing precisely the tasks automated. This is a clear case of documented AI substitution—unlike the generic talk of “digital transformation.”

What role do Social and Economic Committees (CSE) play in AI-related restructurings?

CSEs have the legal right to be informed and consulted before any deployment of AI tools that affect working conditions or job organization. They can request an independent expert to assess the real impact of the deployed tools. This right to expertise is often underused when restructuring plans cite technological motives.

Is the analogy between Leontief’s horses and generative AI really valid?

The analogy is helpful to put doomsday scenarios into perspective, but it has its limits. Horses were replaced because their physical work was substitutable; generative AI is, for the first time, competing with qualified cognitive work—a qualitative break from previous automation waves. Economists are still debating whether labor markets will adapt as well to this as they did to past industrial revolutions, partly because the range of tasks affected is so wide.