Maryland Study Finds No AI Hit to Labor Demand in 155 Million U.S. Job Postings
Updated
Updated · thebanner.com · Jul 7
Maryland Study Finds No AI Hit to Labor Demand in 155 Million U.S. Job Postings
1 articles · Updated · thebanner.com · Jul 7
Summary
University of Maryland researchers analyzed 155 million U.S. job postings from 2018 to 2025 and found no empirical evidence that AI is cutting overall labor-market demand or causing widespread job loss.
The study instead points to post-pandemic rightsizing, tariffs, high interest rates, high oil prices and a broader “low hire, low fire” market as the main reasons hiring has slowed.
Entry-level roles still made up nearly 13% of listings in 2025, up from under 12%, though more of those jobs now mention AI skills and the mix of available work is shifting.
Technical AI requirements appeared in more than 1% of jobs by late 2025, up from less than 0.25% when ChatGPT launched in 2022, suggesting AI is reshaping skills demand more than total demand.
Young workers remain under pressure — unemployment for 20- to 24-year-olds hit 9.2% last fall — but outside economists said it is still too early to measure AI’s full effect and remote-work disruptions may matter more.
With AI reshaping entry-level work, are today's graduates being prepared for jobs that don't exist yet?
If AI isn't cutting jobs, why do hundreds of applications from experienced workers still lead nowhere?
AI’s True Effect on US Employment: Sectoral Trends, Entry-Level Challenges, and the Need for Better Data
Overview
Artificial intelligence is often seen as a disruptive force, with headlines warning of massive job losses. However, recent comprehensive labor market data tells a more nuanced story. While AI is rapidly changing how work is done, the overall demand for workers remains strong. Researchers stress the importance of relying on real data instead of isolated stories. Projects like the UMD-LinkUp AI Maps Project provide detailed, real-time insights, showing that the job market is more resilient than many fear. This evidence challenges the common belief that AI will lead to widespread unemployment, highlighting the need for careful analysis and adaptation.