Meta CEO Mark Zuckerberg testified in Los Angeles County Superior Court last week, defending his company against claims that social media harms children.
A 20-year-old plaintiff alleges Instagram and other platforms are intentionally designed to addict young users.
A new Georgia Tech study aims to examine whether TikTok’s recommendation algorithm contributes to unhealthy patterns among adolescents.
As legal battles intensify over whether social media platforms are engineered to hook young users, new academic research is seeking to shed light on how recommendation algorithms shape teen behavior and mental health.
Munmun De Choudhury, a professor in Georgia Tech’s School of Interactive Computing, is leading a multi-institutional team that will audit TikTok’s recommendation algorithm using data from more than 10,000 adolescent users. The four-year project is supported by a $1.7 million grant from the Huo Family Foundation.
Big trial in the background
The study comes amid growing scrutiny of social media companies. In Los Angeles County Superior Court last week, Meta CEO Mark Zuckerberg took the witness stand to defend his company against allegations that platforms like Instagram are designed to foster addiction among young users. Critics argue that algorithm-driven feeds play a central role in keeping teens scrolling for hours at a time.
De Choudhury’s team hopes to better understand what young people are actually exposed to when using TikTok and how those exposures may affect their well-being.
“We hope to learn the different types of negative exposures that young people experience when using TikTok,” De Choudhury said. “This can help us characterize what they’re watching and build computational methods to understand the consumption behaviors of these participants and how they’re affected by the algorithm.”
The project is a collaboration with Amy Orben, a professor at the University of Cambridge, and Homa Hosseinmardi, an assistant professor at UCLA.
Major hurdle
One major hurdle for researchers has been limited access to social media data. Platforms have tightened restrictions on their application programming interfaces (APIs), making it difficult for outside researchers to study user behavior at scale.
“We can’t do the type of studies we did 10 years ago with X (formerly Twitter) because the API is much more restrictive,” De Choudhury said. “There are limited ways to programmatically access people’s data now.”
Instead, the team is relying on data voluntarily shared by young participants. Orben collected TikTok archives from more than 10,000 adolescents in the United Kingdom who consented to provide their personal data in compliance with the European Union’s General Data Protection Regulation (GDPR).
Unlike many prior social media studies that focus on what users post publicly, this research includes watch histories — offering a rare window into passive consumption.
“We don’t understand passive social media consumption very well, so we hope to close that gap and learn what that looks like,” De Choudhury said. “Is what they’re consuming directly related to what they’re posting? How does passive consumption affect young people’s mental health?”
Researchers believe the findings could eventually inform design changes aimed at reducing harmful effects. Adolescence is a particularly important period, De Choudhury noted, because early signs of mental health conditions often emerge during these years.