87% of Layoffs in 2025 Had Nothing to Do With AI — Gartner Data Challenges Both Sides

Everyone has a narrative about AI and jobs. The panic camp says AI is destroying employment. The executive camp says don’t worry. Gartner’s data says both are wrong — and the real story is more structural than either side admits.

What Actually Caused Layoffs in 2025

1H 2025

241K

jobs lost

2H 2025

1.15M

jobs lost (4.7x more)

AI-productivity layoffs (Reduce)<1% → 1%

Flat. AI is not driving mass firings.

Hiring freezes (Restrain)>4% → <5%

Primarily macroeconomic, not AI.

Repositioning (pivot to AI)17% → 7%

Slowed substantially. Past the hype phase?

Not AI-related79% → 87%

The vast majority. Macro and company factors.

Source: Gartner, 2026

What This Actually Means

1. AI IS NOT THE JOB KILLER (YET)

Only 1% of 2H 2025 layoffs were AI-productivity-driven. The “AI is destroying jobs” narrative isn’t supported by the data at scale. Most layoffs are the same economic forces that have always driven them.

2. THE REPOSITIONING WAVE SLOWED

In 1H 2025, 17% of job losses were “commercial pivot to AI” — companies restructuring around AI opportunities. By 2H, that dropped to 7%. Either the repositioning is done, or companies hit the Mutation Map wall: they announced the pivot but couldn’t execute it.

3. THE REAL WAVE IS STILL COMING

This data is from 2025 — before agentic AI went mainstream, before the harness layer matured, before companies like Intercom renamed themselves after their AI agent. The FRED Test from the AI Supercycle predicts that AI automates tasks with high Frequency, Repeatability, Error tolerance, and Decomposability. Most of those tasks haven’t been automated yet — because the harness wasn’t ready. It is now.

The Supercycle timing: AI job displacement is on the medium clock (10-20 years of industry restructuring), not the short clock (5-10 years of hype). The 2025 data shows we’re still in the early phase — AI is being adopted, not yet replacing at scale. But the medium clock is ticking.

Business Engineer

The AI Supercycle — The FRED Test

Which jobs get automated first? The FRED Test scores any task on four variables. The framework for reading whether your role is on the short clock or the medium clock.

Read the AI Supercycle →

The Bottom Line

87% of layoffs in 2025 had nothing to do with AI. The “AI is killing jobs” panic is premature — and so is the “don’t worry” reassurance. The real wave is on the medium clock: coming, structural, and predictable. The companies that repositioned early are already done. The ones that haven’t started yet are the ones the FRED Test catches next.

Source: Gartner

Scroll to Top

Discover more from FourWeekMBA

Subscribe now to keep reading and get access to the full archive.

Continue reading

FourWeekMBA