US Employment Divergence Widens Along AI Substitution Lines

The European Central Bank has documented a stark divergence in US employment patterns between 2019 and 2025, driven by exposure to artificial intelligence substitution risk.

Jobs classified as having high AI substitution risk—including economists and graphic designers—declined on average by more than 4% over the six-year period. By contrast, roles with low substitution risk, such as electricians and high school teachers, increased by 13%.

Shifting Composition of the US Labour Force

This employment divergence has reshaped the overall composition of US employment. The share of low-risk jobs in total employment rose from 23% to 25%, while the share of high-risk jobs fell from 35% to 33% between 2019 and 2025.

Overall, jobs with high AI substitution risk grew by around 15 percentage points less than jobs with low substitution risk during this period.

ChatGPT Effect Accelerating Change

The European Central Bank notes that the impact of AI on employment growth has accelerated since the launch of ChatGPT in late 2022, suggesting the pace of this occupational shift is intensifying.

Wage Growth Unaffected—So Far

Despite these employment shifts, AI substitution risk has had no significant impact on wage growth since 2019 in the United States, according to the ECB analysis.

EU Context: Productivity Without Job Loss in the Short Term

In the European Union, firms adopting AI technologies are experiencing higher productivity gains without the technology replacing labour in the short term, according to ECB findings. Moreover, firms with high levels of AI adoption or AI-related investment are more likely to employ additional staff.

However, AI is negatively affecting employment for specific occupational sub-groups, particularly junior workers in highly exposed occupations, indicating that the technology’s impact varies significantly across career stages and sectors.

Source: European Central Bank