An Apollo analysis of 321 occupations found that wages in jobs highly exposed to AI grew 6.7% more slowly after 2023, with no statistically significant employment effect. The gap was 10.7% in the lowest-paid quartile and absent in the highest.
The first measurable mark AI has left on the labour market is not unemployment. Apollo’s chief economist Torsten Slok says the employment effect so far is insignificant, and that the visible damage is to pay.
His analysis with Sania Edlich found that wages in occupations highly exposed to AI grew 6.7% more slowly after 2023 than in low-exposure work. Employment in those occupations showed no statistically significant change.
The distribution is the part worth sitting with. The gap was 10.7% in the bottom wage quartile, 5.4% in the second and 4.0% in the third, and there was no significant effect at all in the top quartile.
The method is unusual and worth stating. The authors matched 321 occupations to labour statistics data from 2015 to 2025, using the Anthropic Economic Index, which measures observed AI usage from actual model interactions rather than theoretical exposure.
They are candid about the limits. Exposure is measured from one company’s usage data, only 321 of roughly 800 occupations were matched, and their most dramatic figure, a 24.3% gap for service workers, is flagged as a small subsample to be treated with caution.
Other evidence points the other way. US statisticians found a 0.2% fall in jobs across 18 exposed occupations while payrolls overall grew 0.8%, and Goldman Sachs reported faster declines in openings in fields exposed to AI substitution, in a market where new entrants are already being squeezed.
Diane Gherson, formerly IBM’s chief human resources officer, offers a reason the job losses may be hard to see. Companies are quietly hiring fewer people into high-attrition, lower-wage roles rather than announcing layoffs.
She also names an accounting distortion that pushes the same way. Severance can be booked as a one-off restructuring charge that investors discount, while retraining lands in operating expenses every quarter, which makes cutting look better on paper than reskilling.
The counterexample she reaches for is European. Ikea retrained call centre staff as remote interior design advisers after automating much of their previous work, and the resulting service has been widely reported as a business worth around €1.3bn.
Slok’s wider claim is that the economy is getting more dynamic rather than smaller, with business formation at the highest rate on record. He also concedes the productivity payoff is unproven, since margins outside the largest technology companies have not yet risen.
Europe has no equivalent study, which is the gap worth noticing. The wage channel Apollo describes would be invisible in most European labour data, and TNW has already reported what AI is actually doing to jobs here without anyone measuring pay this way.
Get the TNW newsletter
Get the most important tech news in your inbox each week.
Facts Only
* Wages in jobs highly exposed to AI grew 6.7% more slowly after 2023 compared to low-exposure work.
* Employment effects in these occupations showed no statistically significant change.
* The wage gap was 10.7% in the lowest-paid quartile and absent in the highest quartile.
* The distribution gap for wages was 10.7% in the bottom wage quartile, 5.4% in the second quartile, and 4.0% in the third quartile.
* A gap was absent in the top quartile.
* Authors matched 321 occupations to labor statistics from 2015 to 2025 using the Anthropic Economic Index for exposure measurement.
* The most dramatic figure cited is a 24.3% gap for service workers, which is flagged as a small subsample.
* US statisticians found a 0.2% fall in jobs across 18 exposed occupations while payrolls grew by 0.8%.
* Goldman Sachs reported faster declines in openings in fields exposed to AI substitution.
* Ikea retrained call centre staff into remote interior design advisers after automation.
Executive Summary
Wages in occupations highly exposed to AI grew 6.7% more slowly after 2023 compared to low-exposure work, with no statistically significant employment effect observed. The wage gap was 10.7% in the lowest-paid quartile and absent in the highest-paid quartile. Employment rates in these occupations showed no statistically significant change.
The analysis suggests that the measurable impact of AI is on pay rather than unemployment. A distributional gap existed in wages: 10.7% for the bottom quartile, 5.4% for the second quartile, and 4.0% for the third quartile, with no significant effect observed in the top quartile.
Other evidence suggests a different picture regarding job losses; US statisticians noted a 0.2% fall in jobs across 18 exposed occupations while overall payrolls grew by 0.8%. Furthermore, companies may be avoiding layoffs by hiring fewer people into lower-wage roles and utilizing accounting methods that obscure cost changes related to retraining versus severance.
The economic dynamic appears to be shifting toward greater dynamism, with high rates of business formation, though the productivity payoff remains unproven outside major technology firms. The European context offers an alternative, where retraining initiatives resulted in significant service value creation, contrasting with the measurement gaps observed in US data.
Full Take
The narrative pivots on the distinction between unemployment and wage stagnation as the primary indicators of AI’s labor market effect, which shifts focus from job loss to economic distribution. The statistical finding that employment effects are insignificant suggests that structural changes are being absorbed or masked by corporate actions rather than mass layoffs. However, the observed widening wage gap, particularly in lower quartiles, points to a clear mechanism where productivity gains are not equally distributed across skill levels.
The disparity between measured US results and potential European outcomes highlights methodological dependence on the data framework used—the focus on earnings versus employment is context-dependent based on how labor costs and restructuring events are recorded. The counterpoint provided by Diane Gherson regarding accounting distortions suggests that observed job dynamics may be influenced more by corporate financial reporting choices (severance vs. retraining costs) than by actual labor displacement or creation.
The underlying pattern suggests a decoupling of traditional employment metrics from real economic value generation in the face of AI integration. The challenge lies in recognizing that the observable effects are heavily mediated by corporate strategy and data limitations, leading to an implicit cost borne disproportionately by those at the lower end of the wage spectrum. What forces might cause companies to prioritize delayed restructuring or masking costs over immediate public job announcements? What framework is needed to measure the value of non-modeled human capital investment, like Gherson's retraining example, against short-term financial signaling?
Sentinel — Human
The text reads like a synthesis of specific economic research, skillfully weaving together disparate data points and expert commentary to explore the nuanced impact of AI on labor markets.