Four years after the debut of ChatGPT, the dire predictions of mass unemployment due to artificial intelligence appear to be disconnected from the current reality. According to a survey conducted by Jeffrey C. Dixon, Professor of Sociology at the College of the Holy Cross, the impact of AI on the U.S. labor market is currently limited, though not insignificant.
The Reality of the Labor Market
The study, commissioned through YouGov and surveying 1,250 workers between July 30 and August 4, 2026, reveals a picture far more balanced than what tech pessimists or optimists had predicted:
- Only 3% of workers reported losing a job due to AI since 2023.
- Roughly 6% indicated they landed a job that did not exist before AI (such as AI data labelers or compliance specialists).
- Approximately 9% said they earned a promotion or advancement linked to their AI skills.
While 95% of surveyed workers said "no" when asked if they had lost a job due to AI automation, Dixon notes that these small percentages must be viewed in the general context of the U.S. labor market, where job changes are relatively rare.
The Threat of Agentic AI
While the data currently suggests AI is a ladder for advancement rather than a tool for displacement, the study highlights future uncertainties. The rise of "Agentic AI" — a newer form of technology capable of autonomous action — may render current AI expertise less valuable and automate jobs to a much greater extent than seen today.
"It is difficult for workers to know for sure whether AI is to blame or credit for their job changes," Dixon observes.
The survey was conducted at a time when the official U.S. unemployment rate stood at 4.1%, though the workforce was reportedly shrinking, with AI cited as a leading reason for various layoffs in other sectors.
