Phase I
"The Correction"
Jan 2023 – Dec 2023
Post-ZIRP reckoning. Over-hired during the zero-rate era, tech shed 265K+ jobs. AI displacement was real but marginal — only ~3% of losses were directly AI-attributed. The market was correcting, not yet transforming.
Phase II
"The Reload"
Jan 2024 – Dec 2024
Layoffs normalized at a lower baseline. Companies quietly reorganized around AI capabilities — restructuring teams, cutting manual roles, investing in infrastructure. AI attribution rose to ~13%. The blueprint was being written in silence.
Phase III
"Agentic Execution"
Jan 2025 – Present
The agents arrived. AI-attributed layoffs surged to 45% of monthly totals at peak. Autonomous systems began replacing entire role categories — QA, support, data entry, junior dev. This is no longer correction. It is displacement by design.
751K+Total Tech Layoffs Tracked
3.0%AI Attribution in 2023
13.0%AI Attribution in 2024
22.9%AI Attribution in 2025
Tech Layoffs — Total vs AI-Attributed
Monthly figures, Jan 2023 – Mar 2026 · Shaded regions mark the three displacement phases
AI Displacement Rate (%)
Share of monthly layoffs directly attributed to AI automation · The inflection is unmistakable in mid-2025
The cuts were swift. The consequences, slower. A growing share of companies that displaced workers with AI are now discovering what they lost — and quietly rebuilding what they destroyed.
32.7%Fire-and-Rehire Rate
52.1%Human Oversight Gap
33.0%Skill Loss Impact
20.0%Operational Failure Rate
Regret Metrics — AI Displacement Fallout
% of companies reporting each consequence after AI-driven workforce reductions
32.7%
Fire-and-Rehire Rate
Companies rehiring roles cut for AI within 6 months. The automation savings evaporated before the severance cleared.
52.1%
Human Oversight Gap
HR leaders admit AI requires more human insight than expected. The promise of full autonomy collided with operational reality.
33.0%
Skill Loss Impact
Organizations reporting loss of critical institutional expertise. Some knowledge cannot be documented. It walks out the door and does not come back.
20.0%
Operational Failure
AI tools underperformed or failed to deliver promised savings. One in five deployments did not survive first contact with production.
The per-seat pricing model is collapsing. As AI agents replace human seats, enterprise software companies face an existential crisis.
Q1 2026 Valuation Wipeout
Budget Shift: Seats to Compute