Labor market impacts of AI: A new measure and early evidence

2026

Open publication workspace · Sign in to read the full PDF.

AI-generated summary

This paper introduces a novel measure of AI's labor market impact, combining theoretical capability with actual usage, and presents early evidence suggesting AI is far from its theoretical potential and has not yet significantly impacted overall unemployment.

* Introduces "observed exposure," a new metric for AI's impact on jobs, by integrating LLM capabilities with real-world usage data.
* Analyzes early evidence showing AI's actual adoption is a fraction of its theoretical capability and that occupations with higher observed exposure are projected to grow slower.
* Finds no systematic increase in unemployment for highly exposed workers but suggests a potential slowdown in hiring for younger workers in these roles.

AI, labor market, unemployment, automation, LLM

Check the original publication for accuracy and context.