Artificial intelligence is suppressing wages rather than eliminating jobs—a structural divergence with measurable economic consequences.
Apollo Chief Economist Torsten Slok's analysis of 321 U.S. occupations found that jobs with high AI exposure experienced real wage growth 6.7 percentage points lower than low-exposure roles after 2023, coinciding with widespread adoption of tools like ChatGPT. Employment levels in these exposed occupations remained statistically flat, defying predictions of immediate mass displacement.
The finding isolates AI's primary labor market transmission: downward pressure on worker compensation without proportional job losses. This suggests employers are capturing productivity gains through wage restraint rather than layoffs.
Former IBM HR Chief Diane Gherson said companies are making discrete choices about which tasks to automate and whether AI-driven savings flow to layoffs, hiring freezes, or retraining programs. The divergence between wage and employment effects indicates most firms are choosing the middle path—preserving headcount while moderating pay growth.
A counterbalance exists: Slok noted AI is simultaneously fueling record business formation across sectors, which could generate new job creation and partially offset wage pressure in mature occupations. The net labor market effect remains mixed—weaker compensation growth in exposed roles, stable employment, and an expanding entrepreneurial base.
