Truthout is an indispensable resource for activists, movement leaders and workers everywhere. Please make this work possible with a quick donation.
A new study examining the algorithms of the social media site TikTok appears to suggest that content users received on the platform skewed more politically conservative during the 2024 presidential election, regardless of a user’s personal preferences.
Researchers Talal Rahwan and Yasir Zaki at New York University’s Abu Dhabi campus created over 300 “dummy” bot accounts to mimic real-life users during the 2024 race. The bots “lived” in Democratic-leaning New York, Republican-leaning Texas, and the “purple” state of Georgia.
After creating the accounts, the researchers trained some to follow and react positively to Democratic messaging, while others were trained to support Republican content. Afterward, the researchers monitored the accounts’ “For You” pages for 27 weeks during the campaign, analyzing over 280,000 videos that were suggested by TikTok.
For the Republican bots, the outcome was predictable: the algorithm resulted in an 11.5 percent higher chance that they’d see more Republican-leaning content than Democratic-leaning posts.
But for the Democratic bots, it was different. Those accounts were 7.5 percent more likely to see Republican posts than Democratic ones.
“Our findings show partisan imbalances in political information exposure on a platform dominated by algorithmic recommendations, with implications for platform governance and democratic discourse,” the researchers wrote in their study, which was published in Nature this week.
“The question arises whether the partisan content recommended on TikTok is skewed towards one party over the other,” the authors added.
TikTok responded to the study’s findings, claiming that its conclusions were largely incorrect. In a statement, the company stated:
This artificial experiment with fake accounts does not reflect how people actually use TikTok. In reality, people discover and watch a wide variety of content on our platform which they continuously shape and can control through more than a dozen tools the authors seem unaware of.
But the study’s authors defended their conclusions.
“Our finding isn’t just about reinforcement; Democratic accounts were shown significantly more anti-Democratic content than Republican accounts were shown anti-Republican content,” Rahwan said, speaking to The Guardian about the study’s findings.
“The algorithm wasn’t just giving people what they want; it was giving one side more of what the other side says about them,” Rahwan added.
How social media sites design their algorithms — and whether they are skewed toward one political party over another, accidentally or purposefully — could have real repercussions for elections and the political direction of the country. Indeed, according to data collected from Pew Research, 57 percent of Americans used digital devices “often” to get their news in 2024, higher than any other media, including television (33 percent) and the radio (11 percent).
“In an environment where margins are thin, systematic differences in the kind of political information recommended to tens of millions of young voters are worth taking seriously,” Zaki said.
Please support nonprofit journalism
Independent media is a pillar of democracy and a powerful force for justice. Please make a tax-deductible one-time or monthly donation to Truthout if you can.