Researchers have found that AI smile-changing filters used in real time could have the power to influence who people are attracted to.

A study led by the University of Glasgow used AI to transform faces in real time to study the social cues that influence the beginnings of romantic attraction.

The study used AI to digitally alter smiles during conversations during video speed-dating sessions using volunteers.

During each appointment, the research team used AI to align or misalign participants’ smiles, increasing or decreasing their smiles in real time.

None of the volunteers knew their smiles were being manipulated during the experiment.

The filters used were similar to those already used by social media apps such as Instagram and TikTok, but the researchers said they were more realistic and imperceptible to participants.

The researchers found that even though participants were unaware of their facial transformation, the alignment of participants’ smiles strengthened the romantic attraction they felt toward each other compared to scenarios in which they were not aligned.

They also noted that even when participants did not see their own faces as they transformed, their facial transformations still affected their own attraction rates through feedback loops.

The effects of AI-based smiles also impacted participants’ vocal behavior and their desire to mirror each other’s facial expressions.

Participants also thought the other person was more attracted to them when the researchers increased their partner’s smiles.

Following the experiment, participants were informed that their faces were being altered by the filters, and the researchers explained how this might have influenced any transformations they might have had in their interactions.

The study team says this research highlights a number of key scientific and ethical questions about the potential use of real-time AI facial filters during social interactions, and what implications they might have for research Psychological and ethical studies, particularly regarding facial filters available for public use, are becoming even more sophisticated.

Dr Pablo Arias-Sarah, lead author of the study from the School of Psychology and Neuroscience at the University of Glasgow, said: “This study highlights the potential impact that facial transformation filters, when meticulously calibrated, can have on interpersonal communication.

“From a scientific point of view, this allows us to discover the social parameters that can modulate the emergence of attraction between two potential romantic partners and, more generally, to study how specific social factors influence the way in which individuals interact with each other.

“From an ethical point of view, this study allows us to raise interesting ethical questions. In the coming years, it seems inevitable that AI-based facial and voice filters will become more realistic and prevalent on social media.

“Yet how these technologies might influence users’ social behavior remains largely unknown. As has been done recently for other new technologies, for example deep fakes, a clear ethical guideline on the use and regulation of these technologies may be necessary.

The study was published in the Proceedings of the National Academy of Sciences.