AI isn’t coming for your job — it’s coming for your paycheck.
That’s the conclusion of a new white paper from Apollo Global Management, which measured the observed impacts of AI adoption on wage and employment across various jobs in the US.
The paper, which tracked wage and employment data for 321 occupations, found that jobs with the highest exposure to AI saw an average 6.7% decline in real wage growth after 2023, around the time ChatGPT first went viral.
Yet, AI’s effects on overall employment were not “detectable,” Sania Edlich, an analyst, and Torsten Sløk, the chief economist at Apollo, wrote in a report. Put differently, the takeaway is that AI won’t steal your job but it might cost you a raise.
AI cut most into the wage growth of the lowest earners, the paper added. Service workers saw an average 24.3% decline in earnings growth since 2023, while workers in the bottom 25% of earners saw wages decline by 10.7% over that period. There was no “significant effect” observed among the highest-paid workers, the paper said.
The paper said it examined occupational and wage data from the Bureau of Labor Statistics and acknowledged several “challenges” in analyzing that data over time, such as changes in how different jobs and industries may be classified, and how data is collected. It used Anthropic’s Economic Index, which gauges which jobs are the most exposed to AI, given the percentage of related tasks that have been observed to be performed using Anthropic’s AI tools.
Here are some of the occupations with the highest exposure to AI that have already seen large changes to real wages, per Apollo’s comparison of BLS data from 2022 to 2024:
Some roles saw larger wage declines, which appeared to be triggered by broader industry factors. Broadcaster announcers and radio DJs, who had a relatively low exposure to AI per Anthropic’s Index, saw real wages crater by 52% over the two-year period.
Additionally, some occupations saw a large increase in real wages, despite moderate exposure to AI. For instance, personal finance advisors, whose roles have more than a third of tasks that are exposed to AI, saw wages grow 8.4%. Administrative law judges, adjudicators, and hearing officers, whose roles have 30% of tasks exposed to AI, saw wages surge 17.5% over the same timeframe.
Around 5.8 million workers have roles that are highly exposed to AI, the paper estimated.
“As AI adoption deepens across corporate America, this figure is likely to grow substantially, with significant implications for income inequality and labor market policy in the years ahead,” Edlich and Slok said.
The idea that AI will destroy jobs has faced more pushback this year, but the evidence is growing that widespread use of the technology could cut into some workers’ income, particularly as the labor market evolves to prioritize different skills.
Ioana Marinescu, a UPenn economist, previously told Business Insider that wages could begin to take a hit once roles see around 37% of intelligence tasks being automated.
Historically, workers who have been displaced from “technology-disrupted occupations” took an average real pay cut of around 3% compared to other workers upon finding new employment, and saw real earnings grow an average of 10 percentage points less over the next decade, one Goldman Sachs analysis found.
