Social Media Algorithms: Not the Misinformation Culprits They’re Made Out to Be

New research challenges the prevailing narrative that social media algorithms are primarily responsible for spreading misinformation and extremist content online. A comprehensive study published in Nature by researchers from the University of Pennsylvania’s Computational Social Science Lab reveals that exposure to harmful content on platforms like Facebook is minimal for most users.

Led by Stevens University Professor Duncan Watts, the research team reviewed years of behavioral science studies and concluded that only a small fraction of users encounter false and radical content regularly. More significantly, they found that personal preference, not algorithms, is the driving force behind such exposure.

“The people who are exposed to false and radical content are those who seek it out,” explains David Rothschild of Microsoft Research, one of the study’s co-authors. This finding directly contradicts widely-held beliefs exemplified by statements in major media outlets like The New York Times, which has claimed it is “well known that social media amplifies misinformation.”

The researchers point to how statistics about social media misinformation are often presented without proper context, leading to misconceptions about its prevalence. For instance, while Facebook reported that content from Russian trolls reached approximately 126 million Americans before the 2016 presidential election, this actually represented only about 0.004% of what users saw in their news feeds.

“Citing these absolute numbers may contribute to misunderstandings about how much of the content on social media is misinformation,” Rothschild notes. While acknowledging that misinformation can have significant impact even when rare, the researchers caution against drawing overly broad conclusions from limited data.

Contrary to popular belief, the study found that recommendation algorithms typically push users toward more moderate content rather than extremist material. The research indicates that exposure to problematic content is heavily concentrated among a small minority who already hold extreme views.

“It’s easy to assume that algorithms are the key culprit in amplifying fake news or extremist content,” says Rothschild. “But when we looked at the research, we saw time and again that algorithms reflect demand and that demand appears to be a bigger issue than algorithms. Algorithms are designed to keep things as simple and safe as possible.”

The researchers also challenge the notion that social media is responsible for major societal problems like political polarization. While it’s tempting to correlate increased social media usage with negative social trends over the past two decades, empirical evidence does not support blaming platforms for political incivility or deepening polarization.

To improve understanding and discourse around social media’s actual impact, the team offers four key recommendations. First, they suggest measuring exposure to extremist content specifically among fringe users rather than focusing on average consumption patterns. Second, they emphasize the need to reduce demand for false and extremist content, particularly its amplification by mainstream media and political figures.

Their third recommendation calls for increased transparency from platforms and collaborative experiments between academics and industry to identify causal relationships. Finally, they stress the importance of expanding research globally, particularly in regions where content moderation may be more limited.

The researchers acknowledge that social media remains complex and understudied. “Social media use can be harmful and that is something that needs to be further studied,” Rothschild admits. “If we want to understand the true impact of social media on everyday life, we need more data and cooperation from social media platforms.”

As debates about platform regulation continue, this research suggests that focusing solely on algorithms may miss the more fundamental drivers of harmful content consumption online – the users themselves who actively seek it out.