AI Adoption Shows Limited Impact on U.S. Jobs and Wages Through 2025

Measured Labor Market Shift Despite AI Investment Surge

Artificial intelligence has yet to produce the dramatic employment disruption many predicted, according to research published by the European Central Bank. While certain job categories face clear pressure, aggregate U.S. employment levels and wage growth have remained largely stable through 2025 as the economy absorbs AI technology.

The ECB's Economic Bulletin article, released Monday and first reported by Reuters, analyzed employment patterns from 2019 through 2025 to assess AI's labor market effects. Rather than wholesale job elimination, researchers found evidence of gradual reallocation across sectors.

High-Risk Roles Contract as Low-Risk Jobs Expand

Positions the study classified as having high AI substitution risk—including economists and graphic designers—saw employment decline by more than 4% over the six-year period. Jobs deemed low-risk, such as electricians and high school teachers, grew 13% during the same window.

The research quantified the divergence: jobs carrying high substitution risk grew approximately 15 percentage points less than their low-risk counterparts between 2019 and 2025, holding other factors equal.

This shift has altered the composition of the U.S. workforce. High-risk positions dropped from 35% to 33% of total employment, while low-risk jobs increased their share from 23% to 25%. The ECB characterized this as a reshaping rather than a collapse of the labor market.

Wages Remain Stable, But Future Effects Uncertain

The study found no significant relationship between AI substitution risk and wage growth since 2019. However, researchers cautioned that income effects may emerge as AI tools become more sophisticated and labor markets continue adjusting.

Junior staff in highly exposed sectors appear most vulnerable to displacement, the analysis noted, though the research did not detail specific mechanisms or timelines for potential future disruption.

Why it matters

This empirical evidence challenges narratives of imminent mass technological unemployment while confirming that AI is already reshaping which types of work expand or contract. For business leaders, the findings suggest workforce planning should focus on gradual reallocation and reskilling rather than preparing for sudden, economy-wide displacement. The lag between technology adoption and measurable labor market effects also indicates that current data may not capture AI's full long-term impact, particularly as generative capabilities advance.

The research was first reported by Reuters, with analysis by Balazs Koranyi.