Img source: znetwork.org
Earlier this year, many New Hampshire voters received robocalls with a voice that seemed to be President Joe Biden’s, advising them to stay home during the state’s primary election. However, this voice was not actually Biden’s; it was a “deepfake” created using AI technology. The Attorney General’s Office Election Law Unit in New Hampshire traced the origin of the false AI-generated recording to two Texas-based companies, Life Corporation and Lingo Telecom. Following this incident, the Federal Communications Commission (FCC) has banned robocalls featuring AI-generated voices.
Deepfakes are deceptive audio, video, or images created or altered using AI. As this technology becomes more widespread, there is increasing concern that it will be used more frequently to mislead voters. “The political deepfake moment is here. Policymakers must rush to put in place protections or we’re facing electoral chaos,” stated Robert Weissman, president of Public Citizen. “The New Hampshire deepfake is a reminder of the many ways that deepfakes can sow confusion and perpetuate fraud.”
Besides influencing voters’ opinions about candidates, AI can also be employed to undermine election administration by spreading misinformation that reduces voter turnout. The harmful impact of such high-tech deception is expected to be more severe in the South, a region with a history of election-related deception, where voters already encounter significant barriers to participation.
Disinformation-driven voter suppression is not new in the South, but emerging technologies pose additional challenges for already marginalized voters. As communities of color gain political power, right-wing lawmakers have used disinformation to claim widespread election fraud, fostering distrust in the electoral system and pushing policies that suppress votes to their benefit. Many of these tactics are rooted in Jim Crow-era strategies that aimed to disenfranchise people of color through voter intimidation. For instance, poll taxes required Black voters to pay to register, literacy tests forced them to read a passage of the constitution, while “grandfather clauses” exempted whites from such requirements.
New technologies often bring more systematic voter suppression through “racialized disinformation” via mailings, robocalls, and other mass communications. In 1990, North Carolina Senator Jesse Helms’s campaign sent 150,000 postcards with incorrect voting information, targeting Black North Carolinians and threatening arrests for voter fraud. More recently, Russian bots directly targeted Black voters in 2016, spreading disinformation online to suppress voter turnout in Black communities, according to a report by Deen Freelon from the University of North Carolina. This report indicated that false social media accounts posed as Black users during the 2016 presidential election, exploiting racial tensions to discourage voting in Black communities.
In 2018, social media accounts spread misleading voting information, including false directions to vote by text and claims that voters of one party should vote the day after Election Day. During the 2020 presidential election, there were widespread incidents of digital disinformation. For example, on election day in Texas’s presidential primary, robocalls falsely informed some people that voting would occur a day later, attempting to trick voters into arriving at polling places too late to legally cast their ballots.
With AI technology advancing, voting advocates are concerned that inadequate protections against digital disinformation will impact this fall’s election. Mekela Panditharatne, counsel for the Brennan Center’s Elections & Government Program, wrote, “While it remains unclear how much AI will change the face of vote suppression in the 2024 general election, new developments in AI use and capabilities lend fresh urgency to long-standing efforts to abate attempts to subvert elections.”
