What kind of AI-generated content spread ahead of the 2024 election?
The Civic Listening Corps was formed in January 2022 to give communities across the country the tools and training needed to identify and fight the misinformation we see online every day. Our goal is to use the knowledge we gather together to prevent further attempts to undermine our civic discourse.
By subscribing to our reports, sharing this post, or directing your friends and family to civiclistening.org, YOU can make a difference against the spread of online misinformation.
As the 2024 United States elections approached, many people grew concerned about the rise of AI-generated content—such as fake images, videos, and audio recordings—that could spread misinformation and deceive voters.
The Civic Listening Corps took action by monitoring social media platforms for synthetic media related to social and political topics. Our dedicated team collected and archived over 500 pieces of such content, aiming to track and understand potential threats to the election process.
However, despite these worries, the anticipated wave of misleading AI-generated information did not significantly impact the election.
From October through December 2024, a team of Civic Listening Corps volunteers surfaced AI-generated content across major social media platforms. We defined this content as anything created by a computer based on inputs dictated by humans. Volunteers submitted audio that imitates a real person’s voice, videos, and still images. In total, volunteers surfaced more than 500 pieces of suspect content.
Humor & Entertainment
So what did we find? Generally, these posts were not an explicit effort to misinform or dupe users. While many popular posts referenced or impersonated real people, it was usually obvious to the average user that the content was not based on reality. Though the quality of AI-generated images, video, and audio is rapidly increasing, most content surfaced by volunteers contained clues that the content was synthetic. In many cases, the content was simply absurd or impossible. Much content was created for entertainment or humor.
Rapid-Response Content Creation
We also identified significant amounts of content related to current events. AI tools allow users to create new, interesting, high-quality images and videos as events unfold. We observed this rapid response content creation related to several news stories:
Hurricane Helene’s devastation in North Carolina, South Carolina, Florida, and Georgia
Influencers spreading unsubstantiated rumors about the Haitian community in Springfield, Ohio
The New York State Department of Environmental Conservation euthanizing an influencer’s pet squirrel
Donald Trump comments on making Canada the “51st state”
General election-related content
While the 2024 U.S. elections sparked legitimate concerns about the proliferation of AI-generated content and its potential to mislead voters, such issues did not materialize in a significant way. In 2024, users generated and shared synthetic content more for clout, humor, and commentary than to deliberately mislead. But as technologies improve and become more accessible, vigilance and education will be necessary in our ever-changing information environments.