The Role of AI and Deepfakes in the 2024 U.S. Elections
As the 2024 U.S. elections unfolded, artificial intelligence and deepfake technologies emerged as significant, albeit not dominant, factors in shaping voter perceptions and campaign strategies. While their impact was less dramatic than initially feared, these technologies introduced new complexities to the democratic process, prompting responses from regulators, campaigns, and voters alike.
1. Limited but Concerning Incidents
Several high-profile incidents underscored the potential risks of AI and deepfakes in the electoral landscape:
Robocall Impersonations during US Elections 2024:
In January 2024, just days before the New Hampshire presidential primaries, thousands of registered voters received an AI-generated robocall impersonating President Joe Biden[1]. The call, which sounded remarkably like Biden, urged voters to stay away from the polls, falsely claiming that voting in the primary would make them ineligible for the general election in November[1]. This incident highlighted the potential for AI to be used in voter suppression tactics.
Deepfake Imagery and Videos in US Elections 2024
Political campaigns and individuals shared AI-generated images and videos throughout the election cycle. One notable example involved Donald Trump’s campaign sharing AI-generated images of Trump hugging Anthony Fauci[2]. These incidents demonstrated how easily deepfakes could be created and disseminated, potentially misleading voters or shaping public opinion.
Foreign Influence and Disinformation in US Elections 2024:
Russian state actors actively used AI to generate text, images, audio, and video for disinformation campaigns, often focusing on immigration issues in the U.S.[6] In one instance, a Russian disinformation campaign used AI to create a fake video falsely claiming to show a Haitian man casting multiple votes in Georgia[2]. These efforts aimed to undermine trust in the election process and exacerbate existing social divisions.
2. Regulatory and Legal Responses
In response to these incidents, both federal and state governments took steps to limit AI misuse:
Federal Communications Commission Actions:
The FCC swiftly banned AI-generated robocalls and imposed a $6 million fine on the consultant behind the fake Biden call[3]. This action set a precedent for regulatory responses to AI misuse in political contexts.
State-Level Legislation:
By September 2024, at least 20 states had passed regulations against election deepfakes[4]. Delaware became the 20th state to enact AI deepfake election protections when Governor John Carney signed HB 316 into law[4]. This state-level action demonstrated a proactive and bipartisan approach to addressing the issue.
Legal Consequences:
The political consultant responsible for creating the Biden deepfake robocall was indicted on criminal charges[2]. This legal action sent a strong message about the potential consequences of using AI for election interference.
3. Ongoing Concerns
Despite the relatively limited impact in 2024, election officials and experts remained vigilant about AI’s potential to escalate in sophistication and usage:
Preparedness in Battleground States:
Election officials in key battleground states conducted tabletop exercises to prepare for AI-related disruptions[11]. These exercises included scenarios involving deepfake video and voice-cloning technology deployed across social media to dissuade people from voting or disrupt polling places.
Potential Threats Identified:
The Department of Homeland Security warned that AI tools could be used to create fake election records, impersonate election staff, and more convincingly spread false information online[6]. There were also concerns about AI’s potential to overwhelm call centers with fake voter calls and generate convincing deepfakes.
Technological Escalation:
Experts warned that as AI technology improves, the threat landscape will constantly change, requiring ongoing vigilance from officials, tech companies, and voters alike[6].
4. International Influence
While AI’s impact in the U.S. was significant but contained, other nations experienced a stronger influence:
South Asian Political Campaigns:
Countries in South Asia saw widespread use of AI-generated content in political campaigns[6]. This demonstrated how the technology’s impact could vary significantly by region and political context.
Russian and State Actors:
Russian state actors extensively used AI to generate disinformation content, focusing on issues such as U.S. immigration to stoke divisive narratives among American voters[6]. The U.S. Intelligence Community noted that while these tools did not “revolutionize” such operations, they did “improve and accelerate” attempts to influence voters[6].
5. Public Perception
Despite limited direct impact, public concerns about AI’s influence on elections remained high:
Heightened Public Awareness:
A Pew Research Center poll found that more than 3 in 4 Americans believed AI was likely to affect the election outcome[7]. This high level of awareness reflected the public’s growing understanding of AI’s potential impact on democratic processes.
Concerns Over Negative Impacts:
Over half of U.S. adults reported being “extremely or very concerned” about AI’s negative impacts on the election[7]. This concern spanned across party lines, with similar shares of Republicans (56%) and Democrats (58%) expressing high levels of concern[7].
6. Mitigation Strategies
Several strategies were proposed or implemented to counter the potential risks posed by AI in elections:
Social Media Collaboration and Fact-Checking:
Increased collaboration between social media platforms and fact-checking organizations aimed to counter the spread of AI-generated misinformation[6].
AI Detection Tools:
Advanced AI detection tools were implemented within social platforms and media outlets to help identify and flag AI-generated content[6].
Public Education Initiatives:
Campaigns to educate voters on recognizing AI-generated content grew, aiming to build resilience against misleading AI-generated media[11].
Transparency Requirements for Campaigns:
Proposals for political campaigns to disclose AI usage gained traction, fostering accountability and helping voters better understand campaign tactics[6].
7. Looking Ahead
As the 2024 election cycle concluded, experts warned that the crucial hours and days following the election could still see attempts to use AI-generated content to sow chaos or spread misinformation[6]. The rapid evolution of AI technology means that the threat landscape is constantly changing, requiring ongoing vigilance from officials, tech companies, and voters alike.
In conclusion, while the impact of AI and deepfakes on the 2024 U.S. elections was less dramatic than initially feared, their presence added a new layer of complexity to the electoral process. The incidents that occurred served as a warning of the potential for more sophisticated and widespread use of these technologies in future elections. As such, continued efforts to develop detection tools, implement regulations, and educate the public will be crucial in safeguarding the integrity of democratic processes in the years to come.
