Artificial intelligence is rapidly taking over the workplace, reshaping how people do their jobs and stoking persistent fears around automation-induced job loss. According to a poll conducted by Reuters and Ipsos, more than 70 percent of Americans worry that AI will end up “putting too many people out of work permanently.” But a recent labor market analysis by Yale University’s Budget Lab and the Brookings Institution challenges that assumption, claiming that AI isn’t quite having the impact everyone thought it would.
Is AI Responsible for a Struggling U.S. Job Market?
According to a study by Yale and Brookings, generative AI has done very little to affect U.S. workers’ job prospects so far. In fact, industries were already undergoing occupational shifts before the release of ChatGPT in late 2022, suggesting that other factors may have had a larger impact on the American economy than advancements in AI.
Published in October 2025, the study found little evidence that artificial intelligence has fundamentally transformed the workforce or reshaped the economy in the ways people assumed — at least not yet. In fact, the technology’s impact thus far appears to be fairly limited, suggesting that its long-promised disruption may still be on the horizon. Whether that’s reassuring or ominous will depend entirely on how quickly AI progresses, and how effectively we can create new jobs in the meantime.
What Exactly Does the Study Say?
Yale and Brookings examined how AI is affecting the job market from several angles, and came to a surprising conclusion: So far, AI hasn’t had any meaningful effect on the economy. Here’s what they found.
AI Fails to Distinguish Itself From Previous Technologies
Using the November 2022 release of OpenAI’s chatbot ChatGPT as a reference point, the study compared the percentage of workers who changed occupations post-ChatGPT with similar shifts that occurred when computers and the internet first became popular. All told, less than 5 percent of workers had switched jobs 33 months after ChatGPT’s release — compared with about 6 percent following the spread of computers in the 1980s and just over 7 percent after the internet boom of the 1990s.
This means that only a sliver of the workforce has been tangibly affected by the rise of generative AI and other major technologies. In addition, the results capture job changes without noting the cause, so it’s unclear if AI is the key factor here. Even if it is, it’s projected to affect just one percent more workers than the internet did during the same time span.
Industries Were Already Changing Before ChatGPT
Job changes also started ticking upward in 2021 — a year before ChatGPT’s release — and breaking the results down by industry reaffirms this trend. Each table for the information, finance and professional and business sectors reveals job changes after ChatGPT’s launch that don’t vary much from changes earlier in 2022 or 2021. While 14 percent of workers in the information sector had switched occupations 33 months after November 2022, the study claims that this shift can’t be attributed to a single event like ChatGPT’s public launch.
Putting this into perspective, U.S. workers changed occupations at a much greater rate in the 1950s, when America’s post-World War II economic boom was in full swing. Based on this data, only major world events have driven meaningful economic change in the United States, and the arrival of ChatGPT doesn’t qualify as one.
College Graduates May Not Be Losing Jobs to AI
Next, the authors compare college graduates aged 20-24 with those aged 25-34. If AI has had a meaningful impact on recent college graduates’ job opportunities, more of them aged 20-24 should be entering new occupations. Instead, the percentage of graduates changing occupations has largely stayed between 30 and 33 percent for both age groups. Even 44 months after November 2022, only three percent more graduates aged 20-24 had switched occupations than those aged 25-34, reflecting minor differences between the two groups.
ChatGPT Doesn’t Seem to Be Upending the Economy
For good measure, the study includes data from OpenAI and its rival Anthropic to gauge whether workers in roles most exposed to AI have had to change jobs.
Working with OpenAI’s data, the study measures “exposure” as whether ChatGPT can cut in half the time it takes to complete an occupation’s tasks, grouping workers into lowest, middle and highest exposure. The percentage of workers in each group should fluctuate if ChatGPT is impacting their occupations. However, all three groups have stayed almost the same between 2022 and 2025.
Meanwhile, Anthropic’s data focuses on AI usage, splitting workers into those who use the company’s chatbot, Claude, for automation and those who use it for augmentation. Similarly, the percentages of these two groups should vary over time if Claude is causing workers to switch occupations, yet both have remained almost unchanged between 2022 and 2025.
These findings reveal that generative AI has had no noticeable impact on workers’ job prospects, regardless of the degree to which occupations are exposed to AI. As a result, the study claims that fears over AI taking away jobs are “largely speculative” for the time being.
What Else Could Be Affecting the Labor Market Besides AI?
There’s no denying that the U.S. economy is struggling, but a couple of other factors may be playing a bigger role in the downturn than AI.
Trump’s Chaotic Trade and Immigration Tactics
President Donald Trump has sought to give America the upper hand in its trade relations through tariffs. For instance, he enacted tariffs on hardware as part of the AI race with China, while slapping fees on tech companies for hiring foreign labor through the H-1B visa program. The irony is that these actions could discourage trade with the U.S. and steer tech talent elsewhere — a double whammy for the American economy.
Democrats in the Joint Economic Committee have blasted Trump’s tariffs, identifying them as the cause for the “hollowing out of middle class jobs.” At the same time, Republicans are also worried about the impact of tariffs on the U.S. economy, with the Bureau of Labor Statistics reporting the highest unemployment rate of 2025 in August. If this mounting bipartisan backlash signals anything, it’s that Trump’s tariffs aren’t doing the U.S. job market any favors.
Rising Number of College Graduates
Recent college graduates have been feeling the brunt of America’s economic troubles. According to the Federal Reserve Bank of St. Louis, the unemployment rate for those with at least a bachelor’s degree reached its highest point this past August since 2014, with 2020 being an anomaly due to the pandemic. As the value of a college degree comes into question, many are turning to skilled trades instead of chasing a shrinking number of white-collar jobs after graduation.
But can the lack of early-career tech roles be traced back to AI? Sure, certain roles like software engineering are getting hit hard, but the reality is that there are more college graduates with bachelor’s degrees than ever before, according to Statista. Combine this with an economy that’s been stunted by tariffs and strict immigration policies, and recent grads are condemned to fight over a limited pool of jobs, regardless of AI’s involvement.
Why This Study Should Be Taken With a Grain of Salt
Before dismissing anxiety around AI entirely, it’s important to note that this study has a few flaws, as the authors themselves admit. For one, OpenAI’s exposure data isn’t based on actual usage, making its findings a mere “theoretical estimate.” And although Anthropic does include actual usage data, their results for occupations like coders and writers are likely overrepresented. As for the data on AI’s impact on college graduates, the authors confess that their “results should be interpreted with caution particularly given small sample sizes.”
All these issues point to an overarching problem with the study: There simply isn’t enough data to draw surefire conclusions about AI’s economic effects yet. Because “comprehensive usage data is not publicly available,” the authors had to settle for data from OpenAI and Anthropic. This doesn’t discredit the study, but it does mean that it’s impossible to accurately gauge the true extent of AI’s impact on the job market until more in-depth data is provided.
“To accurately measure AI’s impact on the labor force,” the authors continue, “the most important data needed is comprehensive usage data from all the leading AI companies at the individual and enterprise level, including APIs.”
AI-Driven Job Disruption Remains a Real Possibility
Yale and Brookings may be right about AI’s impact being overblown for now, but this doesn’t rule out the possibility of the technology becoming a disruptive force later on. A Pew Research Center survey published in October found that about a fifth of U.S. workers now use AI on the job, while an Indeed report determined that a quarter of jobs shared on the platform within the past year could be “highly transformed” by generative AI. Even if artificial intelligence isn’t a problem at the moment, it could leave a more noticeable impact during the next recession, according to JPMorgan Chase.
These concerns seem reasonable given the growing number of use cases that can emerge from AI agents, AI web browsers, search APIs and AI-first devices. And with the Trump administration pushing hard on AI innovation and rolling back what little federal regulation there was, it may be too soon to dismiss the technology’s economic impact before stronger evidence proves otherwise.
Frequently Asked Questions
Is AI really responsible for job losses in the U.S?
According to a joint study by Yale University and the Brookings Institution, artificial intelligence — particularly generative AI — hasn’t had a noticeable effect on the job market yet. The authors argue that major occupational shifts were already underway before the release of ChatGPT in late 2022, suggesting other factors besides AI may be at play.
What other factors could affect the job market besides AI?
President Donald Trump’s aggressive tariffs and hardline immigration policies could be placing greater pressure on the U.S. economy and limiting job opportunities as a result. Combine this with the fact that there are more college graduates than ever before, and it’s easy to see why many incoming professionals are struggling to find work.
Could AI still disrupt jobs in the future?
AI could very well upend the job market in the future. A Pew Research study found that one in five U.S. workers now use AI on the job, while JPMorgan Chase expects AI will play a major role in the next recession. The study by Yale and Brookings also uses limited data, so its findings of AI’s muted economic effects aren’t conclusive.
