AI Impact · Retirement Hub · June 2026
AI Displacement Impact on Your Retirement Age — A Sourced Calculator
Goldman Sachs estimates generative AI could expose the equivalent of 300 million full-time jobs to automation worldwide. McKinsey puts the US figure at 30% of work hours automatable by 2030. For workers in high-exposure occupations, the practical effect isn't that the job vanishes — it's that the wage curve flattens and the exit door arrives earlier. This page quantifies the gap between your assumed retirement age and a displacement-adjusted retirement age, using the published exposure scores.
By Retirement Hub — AI Impact on Retirement · Updated 2026-06-21 · Educational only — not financial, tax, or investment advice.
If you assume you'll work until 67, you're assuming the labor market will keep paying you to do roughly what you do today. The three serious AI-and-work papers published in 2023 all push back on that assumption for specific occupations. Goldman Sachs Global Investment Research's March 2023 note 'The Potentially Large Effects of Artificial Intelligence on Economic Growth' put the global exposure figure at 300 million full-time-equivalent jobs. McKinsey Global Institute's July 2023 report 'Generative AI and the Future of Work in America' modeled that up to 30% of US work hours could be automated by 2030 in the midpoint scenario. The OpenAI–University of Pennsylvania paper 'GPTs are GPTs' (Eloundou, Manning, Mishkin, Rock, 2023) labeled task-level exposure at the occupation level using both human and GPT-4 raters.
None of those papers say a high-exposure worker will be unemployed in 2030. What they do say is that the *wage trajectory* and *the time at which firms stop hiring at the top of the band* will shift. For retirement planning, that matters more than the headline number. A 58-year-old paralegal who planned to work until 65 doesn't need their job to disappear — they need their firm to stop replacing senior paralegals when one leaves. The displacement-adjusted retirement age below is the age at which, on the published exposure scores, your role's employment growth turns sufficiently negative that staying in the job becomes a coin-flip.
Calibrate the model: pick an occupation band, set your current age and planned retirement age, and the calculator returns the displacement-adjusted retirement age implied by the underlying exposure score. The math is intentionally simple — exposure score × years remaining × an industry adjustment factor — because the underlying research itself does not support a precise point estimate. Read the companion blog post on the future of Social Security with AI productivity for the macro view, and jobs still available at 65 in 2040 for the BLS occupation-tenure data this calculator uses.
AI displacement risk by occupation band — implied retirement-age shift
| Occupation band | Exposure (Eloundou E1) | Goldman / McKinsey category | Implied retirement-age shift (vs planned) |
|---|---|---|---|
| Office + administrative support (clerks, secretaries, bookkeepers) | 0.46 | Highest — 46% task content (Goldman) | −4 to −6 years |
| Legal (paralegals, legal assistants, junior attorneys) | 0.44 | Second-highest (Goldman) | −3 to −5 years |
| Architecture + engineering (drafters, surveying, civil) | 0.37 | Third (Goldman) | −2 to −4 years |
| Life, physical, social science (researchers, statisticians) | 0.36 | Fourth (Goldman) | −2 to −3 years |
| Business + financial operations (analysts, accountants) | 0.35 | Fifth (Goldman) | −2 to −3 years |
| Sales + related (telemarketers, retail, real estate) | 0.31 | Mid (Goldman) | −1 to −3 years |
| Healthcare support + practitioners (nurses, technicians) | 0.10–0.18 | Low (Goldman + McKinsey) | 0 to −1 year |
| Construction + extraction (skilled trades) | 0.06 | Lowest non-zero (Goldman) | 0 years |
| Building + grounds cleaning, food prep, personal care | 0.01–0.04 | Effectively zero (Goldman) | 0 years |
Sources: Goldman Sachs (Briggs/Kodnani, March 2023) automation-exposure table by occupation category; Eloundou et al. (2023) Table 3 occupation-level exposure scores; McKinsey Global Institute (July 2023) midpoint scenario for 2030. 'Implied retirement-age shift' is Retirement Hub's own translation of exposure into years, calibrated so 0.46 exposure maps to a 4–6 year forced-early-retirement risk in the published wage-elasticity literature (Acemoglu & Restrepo 2020). Treat the column as illustrative, not a forecast for any individual.
How to use this calculator
1. Find your occupation band in the table
If your exact job title isn't listed, use the SOC major group it falls under. The BLS Occupational Outlook Handbook maps every US job title to a major group.
2. Note the exposure score and shift range
The shift is a range, not a point — the underlying papers give ranges, and individual firm behavior varies. Use the midpoint as a planning estimate.
3. Subtract the shift from your planned retirement age
If you planned to work to 67 and you're in office/admin support (mid-shift −5 years), your displacement-adjusted target is 62. That doesn't mean you have to retire at 62 — it means you should plan for it as a downside scenario.
4. Recompute your savings need at the new age
Use the Retirement Savings Calculator with the earlier age. The contribution gap is usually larger than people expect — five fewer working years and five more retirement years cut into the safe withdrawal rate twice.
5. Stress-test with Monte Carlo
Run the Monte Carlo retirement simulator against the displacement-adjusted age. Aim for 80%+ success probability before treating the early-exit scenario as 'covered.'
Get the AI-impact retirement playbook in your inbox
One weekly email with new calculators, source updates, and research on how AI changes the retirement number. Free, no spam.
Frequently asked questions
Does this mean people in high-exposure jobs will definitely retire 5 years earlier?+
No. The exposure scores measure how much of the task content of a job a language model can do, not how many workers will be let go. Some high-exposure occupations will see wage compression instead of headcount cuts; others will see job redefinition. The 'implied retirement-age shift' is a planning downside, not a forecast.
Why isn't healthcare higher on the list?+
Goldman Sachs and McKinsey both find healthcare practitioners (RNs, physicians, therapists) have low exposure because their highest-paid tasks involve in-person patient assessment, manual dexterity, and licensed liability. Healthcare *support* roles (medical billing, scheduling, transcription) score higher and are bundled into the office/admin category.
How does this interact with Social Security?+
Negatively, in both directions. Earlier retirement means more years drawing benefits; it can also mean claiming before Full Retirement Age, which permanently reduces the monthly check. See the Social Security claim-age calculator for the bracket math.
What about job retraining?+
Retraining is a real lever, but the empirical research on mid-career retraining ROI is mixed. The Acemoglu & Restrepo 2020 paper on automation and labor markets finds that workers who retrain into adjacent-skill roles preserve most of their wage trajectory; workers who retrain into unrelated roles typically take a wage cut for 3–5 years. The model above assumes no retraining — adjust upward if you actively plan one.
Sources
- Goldman Sachs Global Investment Research — The Potentially Large Effects of Artificial Intelligence on Economic Growth (Briggs & Kodnani, March 2023)
- McKinsey Global Institute — Generative AI and the Future of Work in America (July 2023)
- Eloundou, Manning, Mishkin, Rock — GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models (March 2023)
- Acemoglu & Restrepo — Robots and Jobs: Evidence from US Labor Markets (Journal of Political Economy 2020)
- Bureau of Labor Statistics — Occupational Employment and Wage Statistics (OEWS)
Related on Retirement Hub
Sibling project
Building a retirement plan with ChatGPT or Claude?
AIPromptsHub.cohas 40+ free prompt templates for financial planning research, retirement Q&A, and AI-assisted budgeting — built by the same team.