The latest jobs report showed that U.S. employment growth surpassed expectations in August. However, wage growth has not kept pace with recent inflation figures, prompting economists to consider a critical question: Could artificial intelligence begin to suppress worker pay before it leads to widespread job displacement?
While the August nonfarm payrolls report offers a snapshot, broader government data signals a deceleration in wage growth. The Bureau of Labor Statistics’ Employment Cost Index revealed that inflation-adjusted wages and salaries saw a 0.4 percent year-over-year decrease through June. Moreover, a concerning long-term trend is evident: labor’s share of nonfarm business output or income hit 52.8% in the second quarter of 2026, marking the lowest point since the series began in the first quarter of 1947, according to the BLS productivity report. Some researchers attribute this decline to decades of automation, a trend that AI is poised to accelerate.
It is important, however, to avoid premature conclusions. The rapid wage growth observed during the COVID-19 pandemic was an anomaly, driven by an exceptionally tight labor market. Current wage gains are more aligned with historical norms. Additionally, higher-paying sectors like technology and professional services have recently experienced job cuts, while lower-paying sectors such as hospitality and healthcare have seen employment increases, naturally pulling down the average wage.
Despite these mitigating factors, the current labor market dynamics are shifting focus from stark warnings of mass unemployment, such as those recently voiced by Bill Gates, towards a more nuanced examination of earnings growth trends. While definitive answers remain elusive, the ongoing discourse is increasingly centering on the potential impact of AI on compensation.
**AI Research Enters a New Phase, Examining Pay Compression**
A recent study by Torsten Slok, chief economist at Apollo Global Management, and co-author Sania Edlich, provides early evidence suggesting AI’s contribution to slower wage growth. Their research indicated that workers in occupations identified as highly exposed to AI experienced real wage growth that was 6.7 percentage points lower after 2023 compared to their less-exposed counterparts. Crucially, the study found no statistically significant impact on employment levels. The authors hypothesize that companies may be leveraging AI-driven productivity gains through wage compression rather than workforce reductions.
While these findings are compelling, labor market experts emphasize the limitations of the current data. Ben Zipperer, a senior economist at the Economic Policy Institute, acknowledges that AI could indeed be influencing demand for specific job types, thereby exerting downward pressure on wages. However, he cautions that the Apollo study’s sample size is too small to draw definitive conclusions.
Zipperer further suggests that such research approaches can overstate the negative impacts of AI. He points to the example of software development, a field with high AI exposure. If AI reduces the demand for software developers by making development cheaper, the cost savings are not lost. Instead, they are likely to be reallocated, potentially creating demand for other workers. “That makes the highly exposed jobs look worse by comparison, even though some of that measured loss is just income increases for other workers,” Zipperer explained.
He also notes that recent job losses in the tech sector, stemming from pandemic-era overhiring, contribute to the observed slowdown in both wage and employment growth. This normalization, rather than AI alone, may explain the weaker figures in tech-related roles. “There was a relative slowdown in labor demand for computer programmers and related jobs in the wake of pandemic rehiring that had nothing to do with AI,” Zipperer stated.
Apollo acknowledged that its study accounted for occupational differences and broader labor market trends. However, it also qualified its findings as “early evidence” based on a limited subset of BLS job categories, with only 321 of approximately 800 BLS occupations being usable and only 11 meeting the study’s high-exposure threshold. The authors highlighted the study’s significance not only for its wage effect analysis but also as a demonstration that “AI research has entered a new phase, one in which labor market impacts can be measured from observed adoption rather than predicted from theoretical exposure.”
Daron Acemoglu, a professor of economics at MIT, points to the limited availability of occupation-specific AI models as a factor influencing current data on job impacts. “AI models are still developing and they are not widely adopted for many occupations or tasks yet. So some of the displacement effects, as of now, may be exaggerated,” he observed.
While Acemoglu sees no convincing evidence yet of widespread wage impacts across specific sectors or demographic groups, he acknowledges “mounting evidence that there is some impact on entry-level jobs.” Given the structure of the U.S. economy, he posits that wages may ultimately bear the most visible brunt of AI’s influence. “Ultimately, given that the U.S. labor market is relatively flexible and has a fairly weak social safety net, I expect the impact on wages to be bigger than those on employment,” he concluded.
Acemoglu’s previous research on the impact of automation on wages and employment informs his perspective on AI’s evolving role. “If AI continues to be developed as an automation technology (most notably under the lodestar of AGI), there will be more impacts. Right now we don’t have many easy to use applications relevant for a large number of tasks/industries. Once these are developed, the labor market effects will be multiplied,” Acemoglu predicted.
**Questioning the “AI Exposure” Narrative**
Economists are increasingly concerned that the framing of “AI exposure” itself might be too simplistic. David Autor, a labor economist and head of MIT’s economics department, argues that knowing an occupation is exposed to AI provides little insight into its future employment or wage trajectory. He emphasizes that the more pertinent question, which he has studied extensively, centers on human expertise and whether AI is encroaching on expert or non-expert aspects of a given role.
Autor and MIT AI researcher Neil Thompson examined two historically similar occupations: accounting clerks and inventory clerks. Both were once considered susceptible to obsolescence in the digital age, performing tasks that could be codified and executed by computers. However, their trajectories have diverged significantly.
Their research found that accounting clerks experienced a 39% wage increase over 40 years, despite a 32% decline in employment. In contrast, inventory clerks saw a 13% decrease in wages, accompanied by a remarkable 175% surge in employment. “These occupations faced the same technological force but experienced opposite outcomes,” the co-authors noted. One became more specialized and better compensated, while the other became more accessible to a broader workforce but less lucrative.
This divergence, Autor contends, challenges conventional narratives of automation and AI exposure. “The conventional wisdom is that as tasks are automated, workers in more automated occupations are pushed downward into lower-paid, less expertise-intensive jobs. The reality, as these two occupations illustrate, is more nuanced,” they stated.
While Apollo’s research focused on AI exposure and pay, other analyses have yielded different conclusions. A February analysis by the Dallas Fed found no direct correlation between overall wages and AI exposure, with a significant caveat: younger workers with less experience, or a lower “experience premium,” may indeed be facing wage pressure. “For occupations with a very low experience premium, AI exposure has a more negative effect on wage growth since AI substitutes for both entry and experienced workers. The low experience premium suggests there is not much tacit knowledge required for the occupation, so experienced workers are easily substituted by AI,” the researchers wrote.
This presents a concerning outlook for those entering the workforce. “The current model of white-collar career progression involves taking an entry level job right out of school and doing codifiable tasks while slowly learning the tacit knowledge to become an experienced worker. Firms are going to find that AI is making this method of employee development cost-ineffective, at least in the short run,” the Dallas Fed concluded. “Of course, leaving new employees off the job ladder is not sustainable in the long run. In the long run, AI adoption will require rethinking how entry-level employees gain experience on the job.”
**Reframing the Human vs. Machine Dynamic**
Ultimately, this discussion points to a shift Acemoglu hopes to see: moving away from an adversarial framing of AI versus human workers. He fears that failing to do so will hinder the development of what he terms “pro-worker AI.”
“The most important thing is that AI does not need to be a pure automation technology,” Acemoglu asserted, adding that AI can also foster new tasks and expertise for workers. However, he conceded that currently, “that’s not the direction we are heading in.” A move towards pro-worker AI would necessitate “the right investments from the tech sector and the right policy framework,” which could foster “better outcomes than this ceaseless race to replace workers.”
Jennifer Huddleston, a senior fellow in technology policy at the Cato Institute, sees positive indicators from both government and industry. She highlighted the Department of Labor’s initiatives to promote AI education within the workforce, including accessible resources for workers to learn about the technology. “Such an approach is likely to serve by helping workers with such a transition rather than protecting or targeting a certain industry or jobs,” Huddleston commented.
Furthermore, she noted that fears surrounding AI in the workplace often overlook its widespread applications in areas like email summarization and cybersecurity, which, while influencing office culture, do not necessarily lead to massive employment shifts. “One often underappreciated element is the way AI is leading to potentially new categories of jobs and opportunities for entrepreneurship. This can mean AI is creating jobs through such new opportunities even if the jobs are not directly related to AI itself,” Huddleston added.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/25679.html
