Despite widespread concern that artificial intelligence will eliminate jobs, new research reveals a surprising trend: occupations most vulnerable to AI automation are not only growing but are doing so at a faster rate than before the pandemic. This counterintuitive finding challenges the prevailing narrative of imminent, large-scale job displacement caused by technology.
An analysis from Vanguard indicates that both employment and real wages in roles with high AI exposure have accelerated in the post-Covid era, outpacing growth in less-exposed sectors. The data suggests that while AI is beginning to reshape workflows, its impact on the broader labor market remains limited, at least for now.
Key Takeaways
- Jobs with high exposure to AI automation saw employment grow by 1.7% in the post-Covid period, up from 1% pre-Covid.
- Real wage growth for these same jobs accelerated from 0.1% to 3.8%, significantly outpacing other occupations.
- Despite anecdotal reports of AI-driven layoffs, widespread evidence of job destruction has not yet materialized in national data.
- Current limitations of AI, such as reliability and accuracy issues, are seen as a key reason for the slower-than-expected disruption.
An Unexpected Boom in Vulnerable Fields
Recent data paints a picture that directly contradicts the fears of many workers and predictions from some industry leaders. A study focusing on approximately 140 occupations identified as highly susceptible to AI automation found that these roles are experiencing a period of robust growth.
Employment in these fields, which include positions like office clerks, HR assistants, and data scientists, increased by 1.7% between mid-2023 and mid-2025. This marks a significant acceleration compared to the 1% growth rate observed in the pre-Covid period of 2015 to 2019. In contrast, job growth for all other occupations has slowed down over the same period.
Adam Schickling, a senior economist at Vanguard who worked on the analysis, noted the clarity of the findings. “At a high level, we have not seen evidence that AI-exposed roles are experiencing lower employment,” he stated, emphasizing that the data does not support the idea of widespread damage from the technology at this stage.
Wages Tell a Similar Story
The trend extends beyond just job numbers and into compensation. The analysis found that real wage growth, which is adjusted for inflation, has surged for workers in AI-exposed professions.
Wage Growth Comparison
AI-Exposed Jobs: Real wage growth jumped from just 0.1% pre-Covid to 3.8% post-Covid.
All Other Jobs: Real wage growth saw a much smaller increase, moving from 0.5% to 0.7% in the same timeframe.
This acceleration in pay is particularly significant. If AI were systematically replacing human labor or devaluing skills, economic models would predict downward pressure on wages. The current data shows the opposite is happening, suggesting that demand for skilled workers in these roles remains strong.
The Gap Between Data and Perception
The optimistic data stands in stark contrast to warnings from some prominent figures in the tech industry and anecdotal evidence from businesses. In May, Anthropic CEO Dario Amodei expressed concern that AI could eventually eliminate a large portion of entry-level white-collar jobs, potentially pushing unemployment rates as high as 20%.
“It’s eerie the extent to which the broader public and politicians, legislators, I don’t think, are fully aware of what’s going on,” Amodei said in an interview. “We have to act now. We can’t just sleepwalk into it.”
These fears are not entirely unfounded. Some business reports reflect early signs of AI-driven workforce changes. The Federal Reserve's November Beige Book, a regular summary of economic conditions, included accounts from businesses beginning to use AI to reduce hiring.
What is the Beige Book?
The Beige Book is a report published by the Federal Reserve eight times per year. It gathers anecdotal information on current economic conditions from business contacts, economists, and other sources in each of the 12 Federal Reserve Districts. It provides a qualitative, on-the-ground supplement to quantitative economic data.
One manufacturer reported cutting its office staff by 15% by implementing AI tools. Another business noted that even small deployments of AI could allow them to skip hiring a new class of entry-level workers. These isolated instances, however, have not yet translated into a measurable negative trend across the entire economy.
Focus on Entry-Level Workers
A primary concern has been the impact of AI on entry-level positions, which often involve routine tasks that are ripe for automation. If companies use AI to handle these duties, it could become harder for young people to start their careers.
However, the Vanguard analysis, which reviewed internal data from 5 million 401(k) plan participants, did not find evidence of this trend. The proportion of workers aged 21 to 25 enrolling in retirement plans has remained stable, suggesting that hiring for this demographic in AI-exposed fields has not declined.
Some tech leaders argue that forgoing entry-level hires is a shortsighted strategy. Cisco President Jeetu Patel called it the “stupidest thing a company can do” in the long run, emphasizing the importance of developing future talent.
“I reject the notion that humans are going to be obsolete in like five years, that we’re not going to have anything to do and we’re going to be sitting on the beach,” Patel said at an AI conference.
Why Hasn't a Job Apocalypse Happened Yet?
The primary reason AI has not yet caused mass job losses appears to be its current technological limitations. While impressive, today's AI models are not yet reliable enough to fully automate complex professional roles without human oversight.
Issues like “hallucinations,” where AI generates incorrect or nonsensical information, remain a significant barrier to widespread adoption for critical tasks. “I am perpetually surprised and amazed with AI’s capabilities – but also with how wrong these models can be at times,” Schickling explained. “It’s clear AI still has limitations.”
The future, however, remains uncertain. The pace of AI development is extraordinarily fast, and improvements in reliability and capability could change the landscape quickly. Vanguard's own analysis predicts that professions like customer service, paralegals, and even economists will likely see lower demand for human workers in the long term.
For now, the data suggests a period of transition rather than immediate replacement. Workers and companies are adapting to new tools that augment their productivity, leading to a demand for skills that can leverage AI effectively, which may partly explain the concurrent growth in jobs and wages.