The Threat of Deepfakes in the 2024 Elections

As the 2024 elections approach, a growing concern is the use of deepfakes to influence the outcome. Deepfakes are AI-generated media, such as videos, audio, or images, that appear realistic but are fabricated. This technology has advanced significantly, making it easier to create convincing deepfakes.

Recent Incidents

• In January 2024, a deepfake robocall using President Biden’s voice attempted to discourage New Hampshire voters from participating in the primary election.
• In 2023, deepfake technology was used to clone a Chicago mayoral candidate’s voice on a fake news outlet on Twitter.

Increasing Sophistication and Accessibility

• AI tools for creating synthetic media have gone mainstream, allowing for the imitation of real people’s voices and appearances.
• The technology to create convincing deepfakes has advanced, making it easier, faster, and more affordable to produce realistic fake media.

Potential Threats

• Deepfakes can be used to spread false information about candidates, manipulate political landscapes, and undermine trust in electoral processes.
• They can be employed for personalized targeting, such as providing false information about polling places to specific voters.
• AI-generated content can be used to create false “evidence” of candidate misconduct, fake videos of election workers destroying ballots, or phony emails directing voters to incorrect polling locations.

Global Concern

• Experts observe a global uptick in electoral deepfakes, with the World Economic Forum’s Global Risks Report warning of the potential for unrest, including violent protests, hate crimes, civil confrontation, and terrorism.

Regulatory Responses

• Lawmakers in 27 states introduced bills to regulate deepfakes in elections within the first six weeks of 2023.
• The FCC has ruled that robocalls using AI-generated voices are now illegal under federal telecoms law.

Tech Company Actions

• Companies like Meta, YouTube, and TikTok now require users to disclose when they post AI content.
• Google and Microsoft have announced measures to protect electoral processes from deepfakes.

Challenges in Combating Deepfakes

• The difficulty in quickly investigating and debunking deepfakes poses a significant challenge, especially in the critical period leading up to an election.
• The cost of production and dissemination of deepfakes is minimal, and they can be spread virally with minimal effort.

Potential Impact on Voter Behavior

• While deepfakes are concerning, some experts suggest that their impact may be limited due to hardened partisan loyalties and the return to pre-COVID social connections.
• However, in close elections, even a small number of people believing false information could potentially affect the outcome.

Broader Implications

• Deepfakes are not just a political issue; they have been used for harassment and abuse, particularly targeting women and potentially limiting their political participation.
• The threat of deepfakes extends beyond elections, potentially undermining trust in media and institutions more broadly.