AI Job Losses Could Cost Workers a Decade of Wages, Goldman Sachs Warns

The next wave of AI layoffs may not just cost workers their jobs — it could cost them a decade of earnings. A new report from Goldman Sachs economists, covered this week by The Wall Street Journal, warns that workers displaced by artificial intelligence could face prolonged wage setbacks that mirror the long, painful recoveries seen after past technology shocks like textile mechanization and manufacturing automation. For job seekers tracking this shift, Metaintro is following how each new round of AI restructuring is reshaping who gets hired, who gets cut, and how long it takes to bounce back.

The headline finding is sobering: when entire occupations disappear, workers don't simply slide into a new role at the same pay. They take pay cuts, churn through lower-quality jobs, and in many cases never fully recover their pre-displacement earnings — even ten years later. Goldman's economists argue the AI transition is now showing the same early signals that defined those earlier waves, and the workers most exposed are the ones with the least cushion to absorb a long setback.

Why does Goldman Sachs think AI displacement looks like past technology shocks?

Goldman's analysis leans on decades of labor research showing that technology-driven job loss is qualitatively different from a normal layoff. When a factory closes because of a recession, workers can often return to similar roles when the cycle turns. When a factory closes because the work itself has been automated, those jobs never come back — and the skills built around them lose market value almost overnight.

The report draws direct parallels to two historical episodes. The first is the mechanization of textile manufacturing in the late 20th century, which hollowed out mill towns across the U.S. South and the U.K. Midlands. Studies tracked by the Bureau of Labor Statistics and academic economists found that displaced textile workers saw earnings losses of 15% to 25% that persisted for more than a decade. Many never matched their previous wages, and entire regional labor markets took a generation to stabilize.

The second parallel is the manufacturing automation wave that accelerated in the 1980s and 1990s. Research by economists including David Autor at MIT has shown that workers displaced by industrial robots and computerized production lines experienced sharp, long-lasting drops in lifetime earnings — particularly men in mid-career roles who lacked college degrees. Goldman's economists argue the AI wave could affect a much broader, whiter-collar slice of the workforce, but the recovery curve looks similar.

The takeaway: AI is not just compressing one job category. It is compressing the transition pathways workers normally use to recover. When coding assistants, document review tools, and customer-service bots remove the bottom rungs of multiple career ladders simultaneously, displaced workers have fewer adjacent jobs to step into.

Which workers face the longest recovery and biggest pay cuts?

Goldman's report flags several groups as especially vulnerable. The common thread is not education level but task substitutability — how easily a worker's day-to-day responsibilities can be replicated by a generative AI system.

Administrative and clerical staff sit at the top of the risk list. Roles in scheduling, data entry, basic bookkeeping, and document preparation are exactly the kind of structured, text-heavy work that large language models handle competently. The BLS already projects declines in secretarial and administrative assistant employment through 2033, and Goldman's analysis suggests displaced workers in these categories face some of the longest re-employment delays because adjacent office roles are also shrinking.

Customer service and call center workers are the next cohort. Companies including Klarna, IBM, and several major telecoms have publicly tied workforce reductions to AI assistant rollouts over the past two years. For workers who built careers in contact centers — often without four-year degrees but with valuable soft skills — the question is where those skills get rewarded next. Retail and hospitality absorb some, but typically at lower pay.

Entry-level legal, finance, and consulting analysts are the third group, and the one drawing the most headlines. Tasks like contract review, due diligence summaries, and first-draft research memos are now handled by AI tools at a fraction of the cost. Goldman's report notes this collapses the traditional apprenticeship pipeline that turned junior analysts into mid-career professionals. Displacement here doesn't just hit current workers — it removes the on-ramp for the next generation.

Mid-career professionals between 45 and 55 face a different problem: age-bias compounded by skill mismatch. Even when willing to retrain, older displaced workers consistently take longer to find new roles and accept steeper pay cuts when they do. Goldman's economists warn this group is the most likely to never recover prior earnings.

What are economists and policymakers recommending to soften the blow?

The report doesn't just diagnose the problem — it lays out a menu of interventions, most of which require coordination between employers, governments, and educational institutions that historically hasn't materialized fast enough.

Wage insurance is one of the most-cited proposals. Programs in Canada and parts of the EU partially compensate displaced workers who take a new job at lower pay, bridging the gap for one to two years while they rebuild skills and seniority. U.S. versions exist in pilot form but reach a tiny fraction of eligible workers. Goldman's economists argue scaling these programs could prevent the long-tail wage losses that define past displacement waves.

Portable benefits — health insurance, retirement contributions, and training credits that move with the worker rather than the employer — are another priority. As AI accelerates job churn, workers who lose employer-sponsored coverage during a transition often delay retraining to chase any available paycheck, locking themselves into lower-wage trajectories.

Reskilling at scale is the third pillar, and the most contested. Short bootcamps and certificate programs have a mixed track record; the workers who benefit most are usually those who needed the least help. Goldman's report points to longer, employer-partnered apprenticeship models — similar to those used in Germany and Switzerland — as more effective for the mid-career cohort most at risk. The catch is cost and time: these programs take 12 to 24 months and require committed employer demand on the other end.

Place-based support rounds out the recommendations. When a single large employer cuts hundreds of AI-replaceable roles in a mid-sized city, the local labor market can absorb the shock for years. Targeted regional investment, infrastructure projects, and small-business support help create the demand that displaced workers need to land softly.

What should job seekers do right now to protect themselves?

Policy fixes take years. Workers reading this today need a plan that works regardless of whether wage insurance ever scales. Three moves stand out from the research.

First, map your tasks, not just your title. The workers who recover fastest from technology shocks are the ones who can articulate transferable skills — judgment, client relationships, complex problem-solving — rather than defending a specific job description. If 60% of your week is structured, repeatable work, that's the slice AI is coming for. Build the other 40% into the center of your professional identity.

Second, stay close to revenue. Across every wave of automation, jobs tied directly to bringing money in the door — sales, business development, customer success, account management — have proven more resilient than support functions. Even within shrinking industries, revenue-adjacent roles tend to be the last cut.

Third, build a parallel income floor. The Goldman report's most uncomfortable implication is that recovery is slow even when it happens. Workers with side income, freelance clients, or savings buffers have far more leverage to wait for the right next role rather than accepting the first one offered. That optionality is what shortens the personal version of the wage-recovery curve.

The history is clear: technology transitions create enormous wealth in aggregate and real, lasting pain for the workers caught at the wrong end of them. The AI wave is unlikely to be different. What can be different is how prepared individual job seekers are when their role lands on the wrong side of the line.

People Also Asked

Q: How long could AI-displaced workers face lower wages?

A: Based on historical comparisons to textile and manufacturing automation, Goldman Sachs economists estimate displaced workers could see depressed earnings for up to a decade, with some never fully recovering their pre-displacement wages. Mid-career workers and those in highly substitutable roles face the longest setbacks.

Q: Which jobs are most at risk from AI displacement?

A: Administrative and clerical roles, customer service positions, and entry-level analyst jobs in legal, finance, and consulting are the highest-risk categories. The common factor is task substitutability — how easily generative AI can replicate the day-to-day work — rather than education level alone.

Q: What can workers do to protect themselves from AI job loss?

A: Focus on transferable skills like judgment and relationship management, move toward revenue-generating roles that are harder to automate, and build a financial buffer through savings or side income. Workers who can wait for the right next role recover faster than those forced to accept the first offer.

Future-proof your career — Metaintro tracks AI's impact on hiring. Sign up free.