CHAPTER 01 · LAYOFFS · 2020–2026
Hiring & firing trend
Jobs were being cut long before anyone blamed AI for it
Six years of layoff data, and for most of that stretch, the pattern has nothing to do with AI. Overhiring, a pandemic hangover, a brutal correction — the numbers track the economy, not the technology. But is that still true in 2026?
The largest layoff total on record has no connection to AI.
In 2021, during the pandemic hiring boom, layoffs fell to just 6,790 worldwide — the lowest figure in years — as companies held onto staff they had over-hired. That over-hiring caught up with them in 2023, when layoffs reached 176,946, a 39% jump over the prior year.
This 2023 figure is the one most often cited as proof of AI-driven job losses. The timeline says otherwise: the spike predates widespread AI adoption. It reflects companies correcting their own over-hiring, not AI replacing workers.
Sources: historic_layoffs.csv (Layoffs.fyi 2020–2025) · tech_layoffs_2026_tracker_ai.csv (Q1 2026)
AI timeline markers (tap the row above): GPT-3 (2020) · ChatGPT launch (2022) · GPT-4 (2023) · EU AI Act in force (2024) — sources →
By 2025, layoffs had fallen to 86,202 — roughly half the 2023 peak. Then, in Q1 2026 alone, they jumped back to 101,146 — and this time, 61% of companies explicitly named AI as the reason. So what's really going on this quarter?
Source: tech_layoffs_2026_tracker_ai.csv · grouped by stated reason for the cuts
Is AI a pretext, or a genuine cause?
On average, the companies investing most heavily in AI are also cutting the deepest. That doesn't prove AI caused any single layoff — stated reasons vary, as shown above — but spending and cuts are clearly moving together. Each bubble represents a company; size reflects AI spending.
Here's the thing about a press release: it costs nothing to say "AI" in it. If AI is really doing the work of the people who got cut, that should show up somewhere harder to fake — in whether students keep enrolling in these fields, and whether companies keep hiring into them.
What actually happened to the workforce these tools were expected to replace?
ICT = Information and Communication Technology · EU27 = the 27 member states of the European Union
The data contradicts the narrative of AI-driven tech contraction.
Source: employment_by_sector.json (Eurostat) · enrollments_by_field.json (Eurostat) · full breakdown next
SECTION 02 · THE AGGREGATE REALITY CHECK
Not just ICT. Not just a good year. The whole labor market.
If AI were reducing European employment, it would appear here first — every sector, every year. It does not. Employment rose across all sectors throughout the years AI tools became mainstream.
EU employment data, across different sectors, 2015–2024
Source: employment_by_sector.json (Eurostat) · EU27, thousands of persons, advertising + education + film & media + health + ICT + R&D combined
Now watch it country by country
The same six sectors, combined per country, shown year by year from 2015 to 2024. By default, bars show sector employment as a share of each country's population (sector employment ÷ population × 100) for direct comparability between small and large countries; select "Numbers" for raw headcounts. Press play to see rankings shift over time.
Source: employment_by_sector_countries.json (Eurostat) · thousands of persons, ages 15–74 · advertising + education + film & media + health + ICT + R&D combined · top 10 countries shown
Student enrollment, the same decade
Source: enrollments_by_field.json (Eurostat) · EU27 tertiary enrollment by field of study, 2015–2024
The enrollment surge predates the release of ChatGPT, accelerating from 2018–2019 — well before AI became a stated factor. If AI were deterring students from technology fields, this trend line would show a downturn. It does not.
Want to check your own country or sector? Explore the full country & sector data →
CHAPTER 03 · SUMMARY
Three charts, one argument.
The full picture: AI and the global workforce
Three years of data across more than 30 countries in the Organisation for Economic Co-operation and Development (OECD). AI has not hollowed out the tech workforce — it has reorganized it. The concern was warranted; the predicted collapse did not materialize.
Simon Blum · SAP FULLSTACK DEVELOPER
"We won't really need first- or second-level support anymore. AI can probably replace that completely — small teams where IT specialists develop alongside the AI and still handle the third-level support themselves."
FIG 04C · INTERACTIVE · ROLES CHANGING WITH AI
Click your field to start, then match roles being replaced to their AI-era successors.
Select a role on the left, then select its replacement on the right. Build all pairs to complete the challenge.
No public dataset maps roles to their AI-era equivalents one-to-one across industries. The mappings below synthesize research from the World Economic Forum's Future of Jobs Report 2025, LinkedIn Emerging Jobs, McKinsey, the Microsoft Work Trend Index, and Gartner, and represent illustrative examples rather than direct predictions. Full data notes →
KEY FINDINGS
What the data actually showed.
One layoff number, four different stories
The scariest layoff number on record predates AI
Wall Street did not read the layoffs as a warning
ICT employment and enrollment both grew, not shrank
It's not just ICT — the whole labor market kept climbing
The transition is uneven across countries
Tom Grünwald · IT SUPPORT ENGINEER
"I don't foresee job losses, but rather a shift in work focus. AI takes over routine tasks — but an experienced person still has to understand, verify, and take responsibility for the results. Control remains with humans."
SUMMARY
Between 2015 and 2024, employment grew across all six tracked sectors — advertising, education, film & media, health, ICT, and R&D — adding roughly 7.1M jobs EU-wide (+21.6%), while enrollment climbed across all ten fields of study, not just technology. ICT led the pack, with EU specialists growing from 6.6M to 10.5M by 2025 (+59%), but the growth wasn't confined to tech.
Tech layoffs peaked at 176,946 in 2023, often attributed to AI-driven displacement. The data suggests otherwise: the peak coincided with companies correcting pandemic-era over-hiring rather than a structural shift toward AI. By 2025, layoffs had fallen by more than half.
CONCLUSION
The data offers a clear answer to "Will AI kill your job?": not yet, and not at the scale suggested by headlines — in tech or anywhere else. AI has accelerated workforce growth across every sector tracked here while placing targeted pressure on automatable sub-roles.