The share of American workers worried about losing their job due to artificial intelligence nearly doubled in just one year, jumping from 5 percent at the end of 2024 to just over 10 percent by the end of 2025, according to a new study from the Federal Reserve Bank of Boston. The research, which surveyed household heads in December 2024 and again in December 2025 through the Federal Reserve Bank of New York's Survey of Consumer Expectations, found that anxiety over AI-related job disruption grew across nearly all industries, educational levels, and age groups. The report concludes that workers increasingly believe AI will reshape their jobs rather than eliminate human work entirely, but the path forward appears uneven and anxiety-laden for many.
The data reveals sharp increases in concern among specific groups and industries. Workers holding doctorates or professional degrees—none of whom expressed job-loss fears in 2024—saw the most dramatic shift: more than 11 percent of professional degree holders and 14 percent of doctorate holders worried about losing their positions in 2025, making them the most likely to anticipate at least one negative job outcome. By industry, consumer services led with 23 percent of respondents concerned about job loss, followed by leisure services at 21 percent and firm services at 15 percent. Meanwhile, 60 percent of all workers in the 2025 survey expected AI-related layoffs or a shrinking workforce in their industry overall, even when they felt their own job was secure. The share of workers expressing concerns about job loss, lower wages, or considering alternative opportunities while saying they couldn't adapt to AI climbed to 18 percent in 2025, up from roughly 11 percent the prior year.
The report identifies a critical dividing line among workers based on how much productivity they've gained from AI. Employees who strongly agreed they'd become more productive since AI's introduction—just 6 percent of the sample—also reported the highest levels of task substitution, meaning they relied on AI for some duties and reallocated their time to other, potentially higher-value work. These highly productive workers felt relatively secure, with only a 6.1 percent estimated likelihood of fearing job loss. By contrast, workers who gave a neutral response to whether AI had boosted their productivity—21 percent of the sample—faced a 21.2 percent estimated likelihood of job-loss concern, the highest of any group. The authors note these workers were likely implementing AI and using it for some tasks but hadn't managed to achieve productivity gains, making them naturally the most fearful for their positions.
The research reveals a mechanism for wage pressure that hinges on perceived productivity. Workers who strongly agreed AI made them more productive had an estimated 14 percent likelihood of saying they were more inclined to ask for a raise in the next 12 months, compared with just 1.9 to 6.4 percent for workers reporting lower productivity gains. The report explains this optimism stems from successfully leveraging the technology to benefit employers, which gives these employees a sense of negotiating power over wages. However, the broader implications for upward wage pressure remain limited because this highly productive group represents only a small slice of the workforce. Workers in the middle—those with neutral or mildly positive productivity experiences—were also the most likely to report facing new, unmanageable work expectations at an estimated 19 percent rate, suggesting they feel they're falling behind at work rather than getting ahead.
The findings point to a labor market in transition, where AI's impact depends heavily on individual workers' ability to harness the technology for tangible gains. The report suggests that instead of wholesale job elimination, AI is enabling workers to spend less time on certain tasks and shift to new assignments, a pattern consistent with recent academic research. Yet this transformation isn't uniform: while some workers feel empowered and ready to negotiate higher pay, a much larger group feels stuck between adopting AI and seeing real benefits from it. The takeaway is clear—AI isn't replacing jobs en masse yet, but it's creating winners and losers fast, and the gap between them is widening.