Losing your job to AI doesn’t just mean finding a new one. According to new Goldman Sachs research, it could mean a decade of reduced earnings — a far longer financial penalty than traditional layoffs impose.
The finding reframes the AI employment conversation from “Will I lose my job?” to something harder to dismiss: “Will I ever recover?”
📉 THE TEN-YEAR PENALTY
Goldman Sachs analyzed labor market data spanning multiple economic cycles and found that workers displaced by automation — specifically AI — face earnings depressions lasting 8 to 10 years. That’s roughly twice the recovery time of workers displaced by typical corporate restructuring or industry shifts.
The mechanism is straightforward but brutal: AI doesn’t just eliminate a position — it devalues the underlying skill set. When a software QA tester is replaced by an AI agent, their testing expertise doesn’t transfer to a new industry at the same wage. They start over, often at a lower pay tier.
This isn’t theoretical. The data is already visible in the current market:
- Entry-level developer hiring for ages 22–25 is down 20% since 2024, per labor market analysis
- AI-specific tech layoffs have surpassed 39,000 in 2026 alone
- Junior engineering roles are being compressed as AI handles code review, testing, and documentation tasks that once served as career on-ramps
🪤 THE ENTRY-LEVEL DISAPPEARANCE
The Goldman analysis highlights a particularly corrosive trend: AI is eating the bottom of the career ladder first.
Entry-level positions — the jobs that train new professionals, build institutional knowledge, and provide the first rung of economic mobility — are the most susceptible to automation. AI agents excel at structured, well-defined tasks. That describes most junior roles precisely.
The result is a generational bottleneck:
- 22–25 year olds face 20% fewer developer positions than two years ago
- Mid-career workers displaced by AI face skill devaluation, not just job loss
- Older workers near retirement see their exit accelerated, shrinking their pension contributions
Senator Mark Warner recently predicted 30% post-grad unemployment if current trends continue — a number that aligns uncomfortably with Goldman’s earnings depression findings.
💰 THE COMPOUNDING DAMAGE
The ten-year earnings hit isn’t just a flat reduction. It compounds:
- Year 1–2: Unemployment or underemployment. Workers take survival jobs at 40–60% of previous salary.
- Year 3–5: Partial recovery as workers reskill, but typically into lower-paying adjacent fields. Each year at reduced salary also means reduced retirement contributions.
- Year 6–10: Earnings may approach pre-displacement levels, but the cumulative wealth gap — missed investments, reduced superannuation, lower Social Security contributions — persists for decades.
A worker earning $90,000 who drops to $50,000 for five years doesn’t just lose $200,000 in income. They lose the compound returns on that $200,000, the career progression they would have had, and the retirement nest egg that never got built.
🔧 WHAT CAN YOU ACTUALLY DO?
Goldman’s data is sobering, but it also points to a clear strategy: act before displacement, not after.
Reskill proactively, not reactively. Workers who begin transitioning to AI-adjacent roles before their current position is automated recover earnings 2–3 years faster than those who wait.
Build human-centric skills. The roles showing the most resilience in Goldman’s analysis are those requiring judgment, relationship management, and physical-world interaction — things AI agents still struggle with.
Negotiate from strength. If you’re in a role likely to be automated, negotiate severance packages that include reskilling benefits while you still have leverage. After displacement, your bargaining power evaporates.
Don’t bank on UBI. While figures like Elon Musk push universal income proposals, Goldman’s analysis implicitly warns that income replacement programs don’t address the skill devaluation at the core of the earnings depression. A check doesn’t restore career trajectory.
🔍 THE BOTTOM LINE
Goldman Sachs has put a number on what many suspected: AI job displacement isn’t just a layoff — it’s a decade-long earnings penalty. The data shows entry-level roles vanishing fastest, mid-career workers facing skill devaluation, and the financial damage compounding well beyond the initial job loss.
If you’re early in your career, the message is urgent: the traditional entry points are narrowing, and the window to pivot is shrinking. If you’re mid-career, the strategy is clear: invest in skills AI can’t replicate before your current role gets automated out from under you.
The ten-year recovery timeline isn’t a prediction — it’s the historical pattern. Breaking it requires acting while you still can.
SOURCES
- Inc. — Goldman Sachs warns that losing your job to AI can hurt your earnings for a decade
