Early Indicators of AI's Impact on Labor: What the Data Is Already Telling Us
For years, economists and technologists debated whether artificial intelligence would disrupt labor markets in meaningful ways — or whether, like previous waves of automation, it would create as many jobs as it displaced. That debate is now shifting from theoretical to empirical. The early data is in, and it paints a nuanced but unmistakably significant picture.
We are no longer speculating about AI's impact on work. We are measuring it.
1. White-Collar Job Postings Are Declining in AI-Exposed Roles
One of the clearest early signals is a measurable decline in job postings for roles that overlap heavily with what large language models (LLMs) can do. Research from economists at the Federal Reserve Bank of St. Louis and independent labor market analysts has found that job postings for:
- Entry-level writing and content roles (copywriters, content writers, technical writers)
- Junior software developers and QA testers
- Data entry and document processing roles
- Customer service agents (voice and text)
- Paralegals and legal research assistants
- Basic financial analysis and reporting roles
...have declined 10–30% year-over-year in the U.S. since 2023, even as overall employment has remained relatively stable. This is not a recession-driven decline — it is specifically concentrated in tasks where AI tools have become functionally capable replacements.
Premium Content
You've read all your free articles today. Subscribe to continue reading.
You've used 3 of 3 free articles today.
Subscribe NowAlready subscribed? Sign in
Comments (0)
Be the first to comment!