Key Takeaways:
As Election Day nears, a widely shared public service announcement on YouTube, now with over 6 million views, is cautioning American voters to scrutinize the content they encounter online, particularly as artificial intelligence (AI) becomes a tool of disinformation.
The video, created by the non-partisan anti-corruption group RepresentUs, uses deepfake technology to simulate the appearance of popular celebrities like Rosario Dawson, Amy Schumer, Chris Rock, and Michael Douglas.
How the Deepfake PSA Was Created
The Don’t Let AI Steal Your Vote campaign leverages deepfakes of familiar faces to drive home its message.
While some celebrities appear in person, others—such as Michael Douglas and Chris Rock—are digitally rendered through AI, emphasizing the deceptive potential of such technology.
RepresentUs CEO Joshua Graham Lynn shared that the campaign had enthusiastic participation from celebrities who wanted to help safeguard democracy.
The Rising Threat of AI in Election Disinformation
The Department of Homeland Security has previously alerted state election officials to the potential of AI tools being used to disrupt election integrity.
RepresentUs CEO Lynn stressed the importance of critical thinking when consuming media, especially with messages that discourage voting or seem unusual.
Experts Warn of the “Liar’s Dividend” and Deepfake Misuse
With the sophistication of deepfake technology growing, disinformation experts are increasingly concerned about its impact on public trust.
Kaylyn Jackson Schiff, a professor at Purdue University and co-director of the Governance and Responsible AI Lab, explains that awareness of deepfakes can lead people to dismiss real events or information as fake, a phenomenon known as the “liar’s dividend.”
Schiff and Ph.D. candidate Christina Walker have been tracking political deepfakes since mid-2023.
Their Political Deepfakes Incidents Database has logged over 500 cases, many of which use deepfakes for satirical or entertainment purposes rather than direct manipulation.
However, even humorous or satirical deepfakes can be problematic, especially if taken out of context.
Identifying Deepfakes: Tips for Voters
Though AI-generated videos are becoming more realistic, they often still contain subtle irregularities that can indicate manipulation.
RepresentUs’s PSA features realistic but not flawless renditions of certain celebrities, with signs of visual editing detectable upon close examination.
As AI advances, spotting these details may become more difficult, but Walker encourages viewers to verify unusual claims through credible channels, especially when election-related content stirs strong emotional reactions. Given the potential for AI-driven misinformation to disrupt elections, experts advise voters to be vigilant: With Election Day approaching and digital tools evolving rapidly, vigilance remains essential. RepresentUs’s PSA serves as a timely reminder of the stakes and calls on voters to stay informed, skeptical, and proactive when consuming information. By recognizing the power of AI to influence narratives and remaining attentive to the sources of their information, voters can guard against the risks of manipulation in a new era of digital technology.
For more news and insights, visit AI News on our website.
Practical Voting Tips Amid Deepfake Concerns
