Bots, deepfakes and fake identities are sabotaging user research

The article, written by the head of research at Mercator Digital, says user research helps product teams find friction early, when fixing problems costs less than later in development. Research cited in the article suggests companies that invest in improved user experience during a project's concept phase cut product development timelines by 33% to 50%.

But that value depends on participants taking part in good faith. The article says disingenuous participation can pollute data and harm the final product. Non-genuine participants are not new; several years ago, the same person attended two focus groups run in close succession for different clients, recruited by different agencies under different names. The difference now is the scale and sophistication of the problem.

Remote recruitment and data collection became essential during the COVID-19 pandemic and remained popular because they reduce travel and logistics costs and let researchers reach participants across wider geographies. The same features that make remote research attractive have also enabled "professional participants" to exploit digital anonymity and low identity-verification barriers for supplemental income. More accessible AI tools have added to the problem, according to the article.

Data from the Market Research Society's Global Data Quality initiative found that 95% of surveyed qualitative researchers reported experiencing participant fraud during the last 12 months. Non-genuine participants can distort study themes, obscure real user needs and lead teams toward conclusions that do not match the target market. A 2025 case study cited in the article reported that usable responses fell from about 75% to around 10% after fraudulent submissions entered the sample.

The article warns against treating every participant who declines to turn on a camera or has audio problems as a fraudster. Users may have legitimate reasons for avoiding video, communicating briefly or having difficulty articulating experiences. Researchers should look for patterns of suspicious behavior rather than rely on a single piece of evidence, it says.

Red flags can appear during application and screening. An unexpected influx of applications from minority groups that are normally difficult to reach may be good luck but should be examined more closely. A study of imposters in online focus groups published in the International Journal of Social Research Methodology noted that receiving a majority of applications from male parents in a study that had previously seen overwhelmingly female participation was an early warning sign. Similar and unusual email addresses, such as common names followed by strings of numbers, are not hard evidence of fraud but can be a useful data point when establishing a pattern.

Another common indicator is inconsistency between what a participant says during screening and what happens in the session. Some claim to use a product every day but then struggle to name basic features or describe a recent interaction. Open-ended questions about recent personal experiences can help reveal discrepancies, as participants may struggle to describe specific events, timelines or terminology associated with the research subject. Some questionable participants appear distracted or disengaged, giving one-word responses, failing to follow questions or resisting attempts to establish rapport. That behavior is not evidence of fraud on its own, because participants can be nervous, tired, uncomfortable or unfamiliar with qualitative research, but it should raise suspicions when it occurs alongside other signifiers.

Editor's Summary User research is increasingly vulnerable to fraud as remote methods, digital anonymity and accessible AI tools enable bots, fake identities and professional participants to enter studies. The TechRadar article cites survey data showing 95% of qualitative researchers encountered participant fraud in a year, and a 2025 case study in which usable responses fell from about 75% to 10%. It recommends pattern-based screening and sensitivity to accessibility needs rather than treating single behaviors as proof of fraud.