In Europe, Women in Tech Decline from 22% to 19% in Two Years

On March 5, 2026, McKinsey published a report on the place of women in tech and AI in Europe. The main finding is summarized in one figure: the share of women in tech roles has fallen from 22% in 2023 to 19% in 2025. In two years, Europe has lost three percentage points of female representation in a sector that promises to redefine the entire economy. [1]

This decline is not a statistical accident. It is the product of three mutually reinforcing dynamics: waves of layoffs in tech, which first affect entry-level positions where women are overrepresented; the rise of AI, which is precisely automating these roles; and the persistence of a glass ceiling that prevents women from accessing technical leadership positions where hiring is concentrated.

Tech Layoffs Hit Entry-Level Positions First

Between 2023 and 2025, the European tech sector experienced a wave of restructuring. Large companies (SAP, Spotify, Klarna, Ericsson) eliminated tens of thousands of jobs, primarily in support functions, product design, digital marketing, and project management. These functions have one thing in common: they are where women had made the most progress in the previous decade.

The McKinsey report shows that women are overrepresented in entry-level positions (product management, UX design, junior data analysis) and underrepresented in senior software development, system architecture, and technical leadership roles. When companies reduce their workforce, they first eliminate the least technical positions, meaning those where women are most numerous. [1]

The result is a pincer effect: women lose jobs in contracting functions and are not numerous enough in expanding functions (AI engineering, cybersecurity, cloud) to compensate.

13% of Tech Management Positions Held by Women

The glass ceiling in European tech has been documented for years. The McKinsey report updates the figures: in 2025, women hold 13% of management positions in tech functions, compared to 30% in non-tech functions within the same companies. The gap is 17 points.

This gap is partly explained by the pipeline: fewer women enter computer engineering programs at university, so fewer women reach senior positions ten or fifteen years later. In Europe, women represent about 20% of computer science graduates, a figure that has been stable for a decade. But the pipeline doesn't explain everything. Women's promotion rates in tech are lower than men's at every career stage, even with equal skills and seniority. [1]

AI Automates Roles Where Women Had Progressed

The irony is cruel. Generative AI, which promises to transform the European economy (McKinsey estimates its potential at €480 billion per year by 2030), is primarily automating tasks in writing, translation, graphic design, data analysis, and content management. These are precisely the areas where women had made the most progress in tech over the last ten years.

Generative AI tools (ChatGPT, Midjourney, GitHub Copilot) do not directly eliminate jobs. However, they increase individual productivity, which reduces the number of people needed to accomplish a given task. A team of five UX writers becomes a team of two, assisted by an AI tool. A team of ten data analysts becomes a team of four. The eliminated positions are those that were most accessible to women.

At the same time, the roles being created around AI (machine learning engineers, AI researchers, system architects) are among the most male-dominated in the sector. In Europe, women represent about 16% of AI professionals, a figure lower than the tech average. [1]

France Within the European Average

France is not escaping the trend. According to data from La Grande École du Numérique and France Digitale, women represent about 28% of employees in the French digital sector in 2025, but only 16% of technical positions (development, infrastructure, data science). The proportion of women in computer engineering training in France has stagnated around 15% for the past five years.

The "Femmes et numérique" (Women and Digital) plan launched by the government in 2023 aimed to increase the share of women in digital professions to 30% by 2027. At the halfway point, indicators show no significant progress. Mentoring and awareness initiatives in middle and high schools exist, but their impact on enrollment in technical fields remains marginal.

What Companies Can Do

The McKinsey report identifies three concrete levers. The first is targeted reskilling: training women in roles threatened by automation in in-demand technical skills (AI engineering, cybersecurity, cloud computing). Several European companies (Siemens, Bosch, SAP) have launched internal retraining programs, but their scale remains limited.

The second lever is revising recruitment criteria. Tech job postings often use language and requirements that discourage female applicants (number of years of experience, exhaustive list of technical skills). Studies show that women apply when they meet 100% of the criteria, men when they meet 60%. Rewording job descriptions and diversifying recruitment channels can broaden the pool of female candidates.

The third lever is pay transparency. The European directive on pay transparency, which is being progressively implemented from 2026, requires companies with more than 100 employees to publish pay gaps between men and women by job category. In tech, where pay gaps are among the highest in the private sector, this transparency could accelerate corrections.

Exceptions That Illuminate the Rule

Some European countries are doing better than the average. Lithuania, Latvia, and Bulgaria show proportions of women in tech higher than 30%, a legacy of Soviet educational systems that encouraged women in scientific fields. Finland, thanks to active parental leave and childcare policies, maintains a rate of 25% women in technical roles.

These examples show that the underrepresentation of women in tech is not a biological or cultural fatality. It is the product of educational choices, public policies, and corporate practices that can be changed. The McKinsey report notes that European companies that have implemented structured mentoring, sponsorship, and work flexibility programs show retention rates for women in technical roles 15 to 20 points higher than companies without specific programs.

The case of Spotify is instructive. The Swedish company published a diversity report in 2025 showing that the share of women in engineering roles had increased from 18% to 26% in three years, thanks to a program combining targeted recruitment, internal mentoring, and diversity objectives integrated into manager evaluations. This result shows that progress is possible, but it requires sustained and measurable investment.

An Economic Issue, Not Just Social

The decline of women in European tech is not just an equality problem. It is an economic problem. Europe lacks tech talent: according to the European Commission, 42% of European companies report difficulties in recruiting digital specialists. Excluding half of the population from the recruitment pool exacerbates this shortage.

McKinsey estimates that closing the gender gap in European tech would add between €260 billion and €600 billion to the EU's GDP by 2030. This figure should be taken with the usual caveats for macroeconomic projections, but the order of magnitude is significant. Europe cannot simultaneously aim to be a global leader in AI and accept that women are increasingly absent from it.

The issue also goes beyond tech. If AI is redefining the skills demanded across the economy, the exclusion of women from the design and development of these tools will have consequences on the very nature of the products and services created. Gender biases in algorithms, documented by researchers like Timnit Gebru and Joy Buolamwini, are partly the product of homogeneous development teams. Diversifying these teams is not just an imperative of fairness. It is a condition for product quality.

The McKinsey report concludes with a sober observation: at the current pace, parity in European tech will not be achieved before 2060. Accelerating requires acting simultaneously on education (increasing the share of women in computer science fields), recruitment (revising criteria and channels), retention (mentoring, flexibility, pay transparency), and promotion (measurable diversity objectives in management). None of these levers is sufficient on its own. All are necessary together.

France: A Revealing Case

France exemplifies the European paradox. The country has a dynamic tech ecosystem (Station F, French Tech, BPI France) and an educational system that produces world-renowned engineers. However, the share of women in computer engineering schools has stagnated around 15% for the past ten years. At École 42, founded by Xavier Niel to democratize access to coding, women represent 14% of students in 2025.

The report from the Haut Conseil à l'Égalité (HCE) published in November 2025 notes that gender stereotypes in educational guidance remain the main obstacle. Girls represent 47% of final-year students in mathematics specialization, but only 16% of those enrolled in scientific preparatory classes (MP and MPI), the tracks leading to computer science schools. The drop-off occurs between high school and higher education, at the precise moment when orientation choices become irreversible.

Initiatives exist. The association Elles Bougent organizes meetings between high school girls and female engineers in 15 regions. The "Femmes du Numérique" program, supported by Syntec Numérique, funds scholarships and mentoring. However, these initiatives remain marginal compared to the scale of the problem. The HCE recommends integrating awareness of digital professions from middle school onwards and setting quantitative diversity targets for higher education computer science programs.

The issue is all the more urgent as generative AI is transforming the labor market at an accelerated pace. Roles in data, machine learning, and prompt engineering are among the most in-demand and best-paid in Europe in 2026. If women remain absent from these fields, the gender pay gap, which has stagnated at an average of 16% in the EU, risks widening rather than narrowing. The AI revolution will be inclusive or it will not be equitable.