New research suggests that YouTube’s recommendation algorithm may shape political views differently for men and women — even if both groups start out with similar interests in political content. The study, published in Cornell University’s arXiv repository, examined how YouTube’s recommendation system responds to different viewing behaviors.
The researchers created 160 automated social bots, dividing them into two groups with “masculine-coded” and “feminine-coded” viewing habits. Although both sets of accounts showed similar interest in YouTube’s News & Politics section, their recommendations reportedly evolved in very different ways over time.
Different algorithms, different political experiences
To conduct the study, the researchers programmed 80 bots with viewing habits associated with traditionally male-oriented content, such as sports and games. Another 80 bots were assigned habits linked to female-oriented content, including fashion, lifestyle, and vlog videos.
Each account then completed 150 consecutive chat sessions, allowing the researchers to monitor how YouTube’s recommendation engine responded.
The results suggested that male-coded accounts tended to focus on controversial and politically charged topics such as crime, law, immigration, and defense issues. These accounts have also reportedly displayed additional content linked to powerful government agencies such as Immigration and Customs Enforcement (ICE) and the Department of Justice.
Accounts coded by women, on the other hand, encountered a broader mix of political content that hinged on international issues, culture, art, and lifestyle-related policy discussions. The researchers also found that these accounts received recommendations for political neutrality in general.
Perhaps most notably, the study found that male-coded profiles were trapped within strong recommendation loops, repeatedly encountering overlapping videos that reinforced the same stereotypes. Accounts with female codes encountered a diverse and less concentrated information ecosystem.
Why the findings matter
YouTube remains the world’s largest content platform and a growing source of news and political information. During the 2020 US election cycle, for example, political campaigns invested heavily in YouTube advertising to influence voters and shape the online narrative.
However, new research shifts the focus away from paid promotions and onto the platform’s recommendation engine itself – the system that determines what users watch next. Jonathan Gray, director of the Center for Digital Culture at King’s College London, said the findings contribute to growing concerns about algorithm-driven political influence and the strengthening of the internet. Gray argued that recommendation systems remain invisible despite their enormous influence on society.
The study also adds to the broader debate about whether large tech platforms are inadvertently increasing alienation by creating personalized echo chambers around users. As scrutiny surrounding AI-driven recommendation systems grows around the world, studies like these may increase pressure on platforms like YouTube to shed more light on how their algorithms influence public discourse and political behavior.
