Artificial Intelligence, Productivity, and the Workforce: Evidence from Corporate Executives

A Federal Reserve Bank of Atlanta working paper analyzes survey responses from nearly 750 corporate executives regarding the impact of artificial intelligence on productivity and the workforce. The authors document substantial heterogeneity in AI adoption, with more than half of firms having invested in the technology, though smaller firms are largely just beginning to adopt it. Labor productivity gains from AI are positive and differ across sectors. These gains are projected to strengthen in 2026, with the largest effects in high-skill services and finance. The improvements stem primarily from rises in revenue-based total factor productivity linked to innovation and demand channels, rather than capital deepening. The study identifies a productivity paradox where perceived gains outpace measured gains, possibly due to lags in revenue realization. There is little evidence of near-term aggregate employment declines from AI. Larger firms anticipate workforce reductions, whereas smaller firms expect modest employment gains. The research also reveals compositional shifts in labor demand. Routine clerical roles are declining, while demand for skilled technical roles is rising. The authors created an index to rank job functions most negatively affected by AI.