AI & Democracy: Does Job Loss Fear Fuel Political Discontent?

VIENNA,Jan. 22, 2026 – Most Europeans believe artificial intelligence is costing jobs, not creating them, and that perception is eroding faith in democracy, according to a new study. The research reveals a surprising link: the stronger people’s belief that AI displaces workers, the less likely they are to trust democratic institutions and participate in political discussions about technology.

AI Job Fears Linked to Democratic Dissatisfaction

A new study shows a widespread belief that AI is destroying more jobs than it creates,leading to decreased trust in democracy.

• A survey of over 37,000 Europeans in 38 countries found a prevailing view that AI leads to job losses.
• Individuals who believe AI replaces human labor exhibit greater dissatisfaction with democratic processes.
• Experiments in the U.S. and U.K. confirmed a causal link: framing AI as job-replacing reduces trust in democratic institutions.
• The narrative surrounding AI substantially impacts public perception and democratic attitudes.
• Targeted communication emphasizing democratic control over AI growth can definately help restore trust.

The research, published in the journal PNAS, demonstrates that these perceptions aren’t merely economic anxieties-they have political ramifications. individuals who believe AI is a job-killer are significantly more dissatisfied with how democracy functions and are less inclined to engage in political discourse or shape technology policy. “When people feel that artificial intelligence replaces human labour, they express doubts about the political system – they are less satisfied with democracy and it’s institutions,” said Christoph Fuchs from the Faculty of Business, Economics and Statistics at the University of Vienna.

What if AI is framed differently? Researchers conducted experiments in the U.K. and U.S. where participants were presented with scenarios depicting AI either as a job creator or a job replacer. The results were clear: when AI was portrayed as replacing jobs, participants showed significantly less trust in democratic institutions.

A Causal Connection

To establish a causal relationship, the research team conducted experiments in the United States and the United Kingdom.Participants were presented with differing future scenarios of artificial intelligence-one emphasizing job replacement and the othre highlighting job creation. The findings revealed that when AI was framed as a threat to employment, participants expressed markedly lower trust in democratic institutions and a reduced willingness to participate in political discussions about AI.

The Power of Narrative

the study underscores that the social consequences of AI aren’t solely dependent on actual job losses. Expectations and fears alone can significantly alter political attitudes. Andreas Raff from the University of Vienna noted, “the very way we talk about artificial intelligence as a society can influence democratic attitudes. If public debates focus heavily on job losses, this can have unintended side effects for democracy.” Fuchs added, “Our findings are a warning signal at a time when democratic legitimacy is declining in many established democracies and when democratic influence on the development of artificial intelligence is critical.”

Shaping the Conversation

The authors emphasize that these effects aren’t immutable. They suggest that targeted communication, highlighting the potential for democratic control over AI’s development and its impact on the workforce, could help counteract the erosion of trust. “Our experiments suggest that people’s beliefs about artificial intelligence are not fixed,” Granulo said. “These beliefs could be changed through targeted communication that highlights that the trajectory of artificial intelligence – and its impact on labour – is not predetermined, but can be shaped through democratic choices.”

The researchers conclude that proactively addressing public perceptions of AI’s economic consequences is crucial for strengthening democratic participation in its development. “If we want to strengthen democratic participation in the development of artificial intelligence, we must take people’s perceptions of its economic consequences seriously – and actively help shape them,” they stated.

Further research is planned through the new Research Network Democracy and Human Rights at the University of Vienna. The university’s semester question for the summer semester 2026 will explore the promises and necessity of democracy and human rights, with contributions available online in the science magazine Rudolphina starting in march 2026.