Urbaniak, from the Institute of Applied Linguistics at the university’s Faculty of Modern Languages initiated and led international research showing that carefully selected words and linguistic structures can effectively manipulate young audiences, persuading them to believe fake news is true.

The study was conducted in Poland and the United States. The Polish findings were published in the journal Scripta Neophilologica Posnaniensia in a paper titled “Gen Z and the crisis of trust: linguistic manipulation in AI-generated fake news.” Professors Erica Grodzki and Martin Phillips of Lynn University in Florida collaborated on the concept and the US research.

The pilot study involved 47 participants in Poland and 47 in the US. Respondents watched news stories presented by an AI-generated avatar. The stories were entirely fabricated by the researchers. After each item, participants rated its credibility on a five-point Likert scale, ranging from “definitely true” to “definitely false.” Follow-up interviews examined why respondents judged the stories as true or false.

Polish Press Agency (PAP): Are members of Generation Z, born roughly between 1995 and 2012 and raised in a fully digital environment, particularly susceptible to linguistic manipulation?

Artur Urbaniak (A.U.): They do show some susceptibility, but the results of this pilot study should be treated with caution. Manipulated messages do affect young people, but not in the straightforward way we initially expected.

PAP: What do you mean by that?

A.U.: The mechanism is complex. What we are seeing is primarily a crisis of trust. Even when faced with potentially absurd information, respondents rarely expressed certainty. Instead of saying something was “definitely false,” they tended to say “it might not be true” or “hard to say.” This suggests they leave open the possibility that false information could be true. In my view, identifying this crisis of trust is the study’s most important finding.

PAP: What does this crisis of trust involve?

A.U.: While respondents were vigilant and in most cases could identify falsehoods, they also questioned true information. In other words, they treated verified facts as potentially false.

PAP: Can you give an example?

A.U.: We presented a clearly false story claiming the Roman Colosseum would be converted into a modern, roofed sports and service facility made of glass and steel. Among Polish respondents, 66% said it was definitely false and 19% said it was probably false. In the US, however, 37% considered it probably true and 9% definitely true. This shows that without solid knowledge or cultural context, people are more likely to make errors.

PAP: What makes young audiences more likely to believe fake news?

A.U.: One key factor is “credentialising” content by using so-called truth markers. These include facts, detailed descriptions, and references to institutions or authorities such as universities, the WHO, UNICEF, or named experts. In the Colosseum example, the story included project details, cost estimates and supposed analyses. Although fake news has been widely studied, truth markers and credibility factors have received relatively little attention from linguists. This is still an emerging field of research.

PAP: Your study focuses on manipulation in text linguistics and the role of LMDs. What does that mean in practice?

A.U.: LMDs are linguistic markers of deception. We wanted to see whether adding them makes false information easier to recognise, or whether more such markers make manipulation more effective. Our preliminary results suggest the relationship is not straightforward. Sometimes these markers increase perceived credibility, sometimes they do not. This indicates the need for further research.

PAP: What kinds of markers of deception did you analyse?

A.U.: Traditionally, researchers identify false information first and then look for linguistic features that characterise it. We instead focused on credibility factors, a specific type of LMD. These included classic rhetorical techniques such as citing false authorities, presenting evidence selectively, creating false dichotomies, and using arguments that appear logical while relying on fabricated data framed in pseudoscientific language.

PAP: What practical contribution can this research make?

A.U.: It can help sensitise young people to fake news in the media. This should not rely on ad hoc measures, such as occasional instructional videos, but on systematic education focused on critical thinking. Curriculum changes are needed to teach rational analysis, questioning and fact-checking.

PAP: How can competencies in critical analysis of media content, especially on social media, be developed?

A.U.: Paradoxically, by strengthening general knowledge about the world, based on reliable and verifiable facts. Young people should learn that information from social media always needs to be questioned. The more knowledge someone has, the harder they are to manipulate. For example, a well-informed person would not believe that average life expectancy in Africa has fallen to under 43 years. Yet in our study, a quarter of respondents thought this might be true, and another 15% were unsure. According to the WHO, life expectancy in Africa is rising and currently stands at 56 years.

The research was carried out under the project “Communicative Competence in the Global Village: Gen Z Facing the Truth, the Post-Truth, and Fake News in Social Media,” funded by the Polish National Science Centre under the Miniatura 5 grant.

Urbaniak has since formed an international research team involving scholars from the United States, Canada, Ireland, Slovakia and Vietnam.

According to Urbaniak, all collaborators consider the project valuable and are willing to continue the work. Due to the lack of domestic funding, he is now considering transferring the research to a foreign centre to continue studies in other languages and cultural contexts.

Anna Mikołajczyk-Kłębek (PAP)

wl/ bar/ mhr/ kap/

tr. RL