The past year has seen promising partnerships in civil society, media, and technology, including AI. express concern About the impact of AI on global elections. Some people may feel scared, but exaggeratedAI has certainly made its presence felt in elections around the world, and it has had a negative impact on voters. As a panel of experts and a PAI partner, our predictions from last year at PAI policy forumthe impact of AI has not been as detrimental as some thought, but the quantity and quality of election content generated by AI has increased.
The ubiquity and accessibility of AI applications has significantly improved the quality of synthetic content, creating confusion, mistrust, and mistrust. desensitization To the flood of generated political content. As the 2024 US election concludes this week, let’s take a look back at the many ways AI contributed to the election, including through the spread of misinformation and deepfake audio, video, and images.
voter suppression
Deepfakes are audio or visual media that is generated or modified by AI. They usually depict a real person saying or doing something that they did not actually say or do. At the beginning of this year, new hampshire voters They were surprised to receive a call from someone they believed to be President Joe Biden urging them not to vote in their state’s primary. The caller pretended to be President Biden’s likeness and identity, but was actually a robocall that used AI to imitate his voice. Adding to the deception, the call was: manipulated to show the caller ID from former New Hampshire Democratic Party Chair Kathy Sullivan. It was later revealed that the two companies, Ringo Telecom and Life Corporation, and political consultant Stephen Kramer were the two companies. The mastermind behind the phone. Since then, Kramer and Lingo Telecom fined by FCC For attempts to deceive and defraud voters. However, Phone calls seemed ineffective In the New Hampshire primary, robocalls served as an example of how bad actors could use AI in the future to influence elections.
make opposing candidates look incompetent
Generated by widely distributed AI videoPresidential candidate Kamala Harris is seen rambling and gibberish during a speech at Howard University. The video was originally shared on X (formerly Twitter) and has racked up over 4 million views. Although this video appears to be authentic online, it was actually generated by AI. Found to be fake by Reutersthe voice quality and noise around Harris’ mouth has been very visibly digitally altered. This video is perceived to be real and may have negatively influenced people’s opinion of Kamala Harris.
Screenshot of Trump’s social media posts displaying fake support for the Trump campaign
fake celebrity endorsement
In August, presidential candidate Donald Trump posted an AI-generated image on his social media platform Truth Social that falsely purported to show Taylor Swift and fans supporting his campaign. Posted. When Swift became aware of the misinformation being spread about her support, she took to social media to complain: “Fighting misinformation… together with the truth.” Recognized as one of the best in the world influential people, swift Publicly supporting Kamala Harris President with an Instagram post. Misuses of AI are prompting lawmakers like Gavin Newsom to take action. Newsom signed in September. invoice Legislation has been enacted to combat deepfake election content and protect the digital likeness of actors and performers. The bill would “remove deceptive content from large online platforms, increase accountability, and improve voter information.”
cater to demographic trends
This election saw the heavy use of AI-generated images on social media to sway voters, especially in undecided communities. image of President Trump poses with black voters It was shared online to garner more support from the black community. Throughout his campaign, Trump specifically hostile The AI images of Black, Indigenous, and people of color were an attempt to repair his reputation among voters in these communities. While unlikely to sway voters who have already made up their minds, these images can reinforce existing beliefs or distort perceptions about particular candidates, potentially swaying undecided voters.
AI-generated image of Donald Trump running away from police officers
make fun of someone
As we observed, deepfakes can take the form of songs. drake vs kendrick feud earlier this year. But during this election cycle, the D.N.C. Generate a deepfake parody Republican National Committee Co-Chair Lara Trump’s heavily auto-tuned truckanything is possible“of parody teeth is advertised as “A summer party anthem about how the RNC is collapsing under Lara Trump and the rest of the new Ultra MAGA team.” The teasing gets more serious when the AI generates an image of Donald Trump. turned in the direction being arrested President Trump’s Face Started Making the rounds on Social Media criminal charges. Trump became the first former president ever to be convicted of a felony, but the fake images from his arrest further damaged his reputation.
Image of a girl holding a puppy generated by AI
spread propaganda
AI-generated images are also being used to spread propaganda online. After Hurricane Helen, an image of a girl holding a puppy on an AI-generated rescue boat was widely shared on X, sparking outrage over the Biden administration’s response to the disaster. The image turned out to be fake, but people felt they had to keep spreading it as a symbol About “the trauma and pain that people are living through right now.” Similar AI images of animals showing signs of support for Trump are circulating online, causing outrage and tugging at the heartstrings of many people. These images remain today. false claims President Trump and his administration have claimed that Haitian immigrants are kidnapping and eating the pets of other Springfield, Ohio residents. Although fake, these AI-generated images are useful tools for spreading misinformation and reinforcing harmful political discourse.
First GenAI election…but not the last
The impact of AI in the 2024 US election may have been minimal, but as AI technology advances and people gain access to AI, it could have a larger impact in future elections .
At PAI, we recognize the risks that AI poses to democracy. Therefore, we will continue to gather experts Addressing these challenges requires collaboration from across civil society, academia, and industry. We remain committed to addressing the challenges that AI poses to democracy and preserving the integrity of our electoral processes. To stay up to date on our work in this area, please visit Newsletter.
