Goldman Sachs AI Job Warnings Collide With Data Showing Rising Wages in AI-Exposed Roles
Wall Street economists predict tens of thousands of monthly job losses from AI, but research from Vanguard, Yale, and MIT shows wages climbing and employment holding steady in AI-exposed occupations.
Goldman Sachs economists warn of 16,000 AI-related job losses each month. Anthropic's chief executive predicts a 50 percent collapse in entry-level white-collar roles. Yet workers in AI-exposed occupations are seeing real wages climb, and the broader labor market remains stable.
Wall Street's apocalyptic alarms clash with data from Vanguard, Yale, and MIT showing AI functions as a productivity multiplier driving wage growth and steady employment. The predicted job apocalypse is not materializing in the numbers. Corporate layoffs appear to be opportunistic cleanups of pandemic-era overhiring rather than genuine technological displacement.
Vanguard research shows real wages in AI-exposed roles grew 3.8 percent from the second quarter of 2023 to the second quarter of 2025. That compares with just 0.7 percent growth in less exposed occupations. Employment growth in high AI-exposure jobs rose from 1 percent before the pandemic to 1.7 percent afterward, while all other jobs declined from 1.1 percent to 0.8 percent.
Overall, our metrics indicate that the broader labor market has not experienced a discernible disruption since ChatGPT's release 33 months ago, the Yale Budget Lab stated in its October 2025 report. The picture of AI's impact on the labor market that emerges from our data is one that largely reflects stability, not major disruption at an economy-wide level.
Venture capitalist Marc Andreessen argues companies are using AI as a silver bullet excuse for cleaning house. He told listeners on the 20VC podcast March 30 that essentially every large company is overstaffed. Most are overstaffed by 50 percent, Andreessen said. A lot of them are overstaffed by 75 percent. Now they all have the silver bullet excuse. It is AI.
This perspective explains recent mass layoffs at Block, which cut 40 percent of its staff. Crypto.com shed 12 percent of workers. Oracle eliminated up to 30,000 positions. With macro-level labor metrics remaining stable, these cuts reflect corporate overhiring rather than systemic technological unemployment.
The AI job loss narratives are all fake, Andreessen posted on X April 6. AI equals a massive ramp in productivity, which equals a massive ramp in demand, which equals a massive jobs boom. Watch.
Goldman Sachs's own research supports this productivity thesis. Management teams quantifying AI impact found median productivity gains of 30 percent in customer support and software development tasks, according to Goldman strategist Ronnie Walker. MIT Sloan research shows firms adopting AI experience approximately 6 percent higher employment growth and 9.5 percent more sales growth over five years.
Goldman economists acknowledge AI has reduced U.S. job growth by about 16,000 net jobs per month. Substitution wipes out 25,000 positions while augmentation adds back 9,000. Yet this net drag represents a relatively small fraction of the productivity gains. The economy is absorbing AI benefits through increased demand and new roles rather than pure displacement.
Workers displaced by technology face real challenges. They take approximately one month longer to find new jobs and suffer earnings losses exceeding 3 percent upon reemployment. Over a 10-year horizon, real earnings growth for tech-displaced workers lags never-displaced workers by nearly 10 percentage points, Goldman research reveals.
Younger, college-educated workers fare better. Their cumulative earnings losses over a decade run roughly half as large as other tech-displaced workers, thanks to occupational mobility and skill upgrading. Workers who participated in vocational programs within three years of displacement saw about 2 percentage points more cumulative wage growth over the following decade.
AI is expected to positively impact about 80 percent of all jobs in the next decade, enhancing job functions rather than replacing jobs entirely, said Joe Davis, Vanguard Global Chief Economist.
The U.S. unemployment rate held steady at 4.3 percent in March, though long-term unemployment rose by 322,000 over the past year. AI adoption reached 17.4 percent of U.S. establishments, according to Census Bureau data, up from 10 percent in late 2023.
Vanguard concluded current AI systems are generally enhancing worker productivity and shifting workers' tasks toward higher-value activities. MIT researchers found that when AI impacts only a few tasks in a job, employment in that role can grow as workers reallocate effort to more valuable work.
The market data shows transformative technologies supercharge worker productivity and value. Current AI anxiety appears less about economic reality and more about political leverage for state interventionism. The free market responds efficiently to the productivity shock, expanding demand and creating new roles faster than displacement rates. Workers adapt, upgrade their skills, and move forward.