Only 3% of U.S. Workers Report Losing Jobs to AI Since 2023

Survey reveals limited AI job disruption

Four years after ChatGPT's November 2022 debut, artificial intelligence has not triggered the widespread job losses many predicted, according to new survey data from U.S. workers.

A YouGov survey of 1,250 employed American adults conducted between July 30 and August 4, 2026, found that only 3% of workers reported losing a job due to AI automation since 2023. Meanwhile, 6% said they landed jobs that didn't exist before AI adoption, and 9% received promotions or advancements they attributed to AI skills.

The findings, first reported by Fortune, come from research led by Jeffrey C. Dixon, a sociology professor at the College of the Holy Cross who studies AI's labor market effects. The results have not yet been peer reviewed.

Most workers report no AI-related job changes

The survey asked three specific questions about AI's role in employment changes since 2023, covering job losses, new AI-created positions, and AI-related promotions. For each question, between 88% and 95% of respondents answered "no," indicating they had not experienced these changes.

Between 3% and 4% of workers said they were "not sure" whether they had experienced AI-related job changes, highlighting the difficulty workers face in determining whether AI directly caused their employment outcomes.

The survey sample was designed to reflect the U.S. employed population across demographics including age, gender, race, and education level.

Why it matters

These findings challenge both pessimistic predictions of mass AI-driven unemployment and optimistic forecasts of widespread job creation. For business leaders navigating AI adoption, the data suggests the technology's labor market impact remains modest in the near term, though the emergence of more autonomous "agentic AI" systems could accelerate changes. The most tangible benefit appears to be career advancement for workers who develop AI skills, though whether this advantage persists remains uncertain.

Context and limitations

Dixon notes the results align with other research showing AI's limited current impact on labor markets. However, the survey captured a specific moment when the official unemployment rate stood at 4.1% in July 2026, and AI was reportedly a leading factor in many layoffs despite the workforce shrinking.

The study differs from much existing research by asking workers directly about their experiences rather than relying solely on economic predictions or company analyses. Unlike some studies focused exclusively on job losses, this survey examined both negative and positive employment outcomes.

Dixon acknowledges limitations: survey participants volunteered for YouGov's online panel rather than being randomly selected, and none were currently jobless. Workers may also struggle to accurately assess whether AI caused their job changes.

The researcher is continuing to analyze the data to answer additional questions about which workers are getting promoted, how they're using AI on the job, and how they expect AI to affect future opportunities.

The findings were published by Fortune and The Conversation under a Creative Commons license, based on Dixon's ongoing research at the College of the Holy Cross.