Global Pew Poll: 46% Predict AI Will Eliminate Jobs

A global survey conducted by the Pew Research Center across 37 nations reveals that public pessimism over artificial intelligence and employment is pervasive. Across a 37-country median, 46% of adults predict that AI will lead to fewer jobs over the next 20 years, compared to just 9% who anticipate job creation and 13% who expect little difference. In 34 of the 37 countries surveyed, “fewer jobs” was the most common response. Anxiety is especially concentrated in wealthier economies, where 71% of U.S. adults and roughly seven-in-ten respondents in Australia and South Korea foresee net job displacement. Public concern significantly outpaces current macroeconomic labor market disruption.

Global poll shows widespread job loss expectations

The comprehensive survey released by the Pew Research Center paints a stark picture of global sentiment regarding generative technologies. Approximately one-quarter of global respondents reported being unsure about long-term labor outcomes, but among those with a firm perspective, skepticism dominated across nearly every surveyed geography, as reported by Gizmodo.

Nigeria was the only nation among the 37 polled where more respondents predicted that AI would create jobs rather than eliminate them, with 27% predicting employment growth versus 26% expecting net losses. In several developing and middle-income markets, uncertainty proved far more common than definitive pessimism. For example, in the Philippines, 54% of adults were unsure about AI’s employment trajectory, while only 26% predicted job cuts. Similarly, in Thailand, 49% answered that they were unsure, compared to 32% anticipating fewer jobs.

Demographic patterns within the survey indicate that anxiety cuts across generational lines, but is felt most acutely among younger cohorts. According to Pew’s demographic findings on young adults, younger generations in both high-income and middle-income countries express heightened concern regarding automation. Concurrently, public apprehension is accelerating across Europe. The share of adults who report feeling more concerned than excited about artificial intelligence rose substantially year-over-year in Sweden (+9 percentage points), Poland (+8), the Netherlands (+7), Hungary (+6), and Australia (+4).

Economic divergence: Wealthier nations report acute labor anxiety

The poll uncovered a pronounced divide based on national economic output. As detailed in Pew’s analysis of high-income countries, populations in advanced economies express vastly more concern over labor replacement than those in emerging economies. In countries with higher gross domestic product, automated cognitive systems directly threaten white-collar, administrative, and knowledge-work roles that form the core of the service economy.

In the United States, 71% of respondents now predict AI will eliminate more jobs than it creates over the next two decades. This reflects a significant increase from 2024, when 64% held that view. Similar levels of apprehension appear in Australia and South Korea, where roughly seven-in-ten adults anticipate employment contractions. By contrast, across middle-income nations, the median share anticipating job cuts stands at 36%.

Educational and income distributions within middle-income markets present an informative dynamic. Higher earners and college-educated individuals were more likely to predict job losses than lower-income peers. Pew noted this discrepancy stemmed not from greater optimism among lower earners, but from higher rates of uncertainty, with lower-income workers predominantly selecting “not sure.”

The survey also highlighted deep ideological divisions regarding economic inequality. In the U.S., respondents identifying with the political left were 31% more likely than conservatives to fear that AI will widen societal inequality. Overall, 46% of Americans believe AI will exacerbate the gap between rich and poor, 21% foresee little impact, and only 5% believe it will make the nation more equitable.

Labor sentiment versus empirical economic evidence

The intense public anxiety captured by Pew stands in sharp contrast to observed macroeconomic data. An empirical analysis published in July 2026 by researchers at the Stanford Institute for Economic Policy Research found that artificial intelligence has had only marginal effects on the broader U.S. labor market to date. Aggregate employment metrics and overall wage distributions have not yet registered the widespread disruptions predicted by automated displacement models.

However, the Stanford researchers identified an emerging vulnerability for entry-level workers. While seasoned professionals continue to leverage generative tools for workflow assistance, recent college graduates face a more constrained hiring environment in sectors adopting automated content generation, basic coding tools, and document processing. This localized friction aligns with the heightened anxiety expressed by younger respondents in the Pew data.

Public pessimism has also been fueled by early corporate messaging. During the initial commercialization waves of generative AI, frontier technology firms frequently marketed enterprise platforms by highlighting payroll reductions and workforce consolidation. Although technology executives have recently pivoted toward framing AI as an assistive productivity tool, the initial emphasis on automation appears to have entrenched widespread public skepticism.

What happens next

The widening divergence between public sentiment and corporate adoption creates distinct operational hurdles for enterprise technology leaders. As organizations integrate autonomous workflows and automated tools, internal deployment teams must account for employee resistance driven by job security concerns. Clear policies establishing whether systems serve to augment existing staff or automate responsibilities will be essential to maintaining internal adoption.

Furthermore, the rising level of public concern in Europe and the United States will likely accelerate regulatory attention. With 71% of American adults expecting negative labor outcomes, political scrutiny over workforce displacement, training initiatives, and corporate restructuring is poised to intensify. Technology vendors and enterprise buyers alike will need to validate productivity gains with transparent operational data rather than aggressive workforce reduction targets.