A new study of Ohio’s 2022 elections concluded social media bots were able to influence what candidates were messaging about online, and the issue has likely only worsened with time, according to its lead researcher.
Bots, or autonomous social media accounts designed to mimic human users, are pervasive on most social media platforms in 2026. Philip Arceneaux, assistant professor of strategic communication at Miami University, said bots aren’t just noise in electoral politics.
“The most significant finding was they drove negative tone or salience of issues across every permutation that we analyzed,” Arceneaux said in an interview.
Led by Miami University and released last Sunday, the study by the Political Public Relations Lab analyzed 900,000 tweets from the race for Ohio governor, the race for U.S. Senate and 15 races for U.S. Congress. Those tweets came from a mix of candidates, journalists and other users.
The campaigns were more likely than the press or the public to be swayed by bots boosting an issue, the study argued.
“They to a certain extent elbowed the campaigns into talking about this more than maybe they wanted to on their own,” Arceneaux said.
With artificial intelligence maturing, he said he thinks it is worse now.
“We don’t know who’s running these bot farms,” Arceneaux said. “We don’t have necessarily the best ideas rising to the top anymore, because they have technical means to break that process, right? It’s not natural selection anymore. It’s artificial selection by who has, who can control the most bots.”
Similar studies of bots' impact on political persuasion have shown similar results. Studies in the journals Nature and Science released last month showed conversations with AI chatbots did influence participants perspectives and led them to change their candidate preference or views—moving them by a couple of percentage points to double-digit differences.
