Around the world, politicians are taking advantage of the new advancements in generative AI to win their election campaigns.
Generative AI uses artificial intelligence to create content such as music, text, audio, videos, and pictures. Midjourney, initially released in 2021, is a popular generative AI program that uses a text-to-image model to design images. OpenAI’s DALL-E is another program that can manipulate and rearrange objects in an image to create photorealistic pictures. These sites, among others, are relatively inexpensive and user-friendly, allowing anyone to create whatever they desire.
Within the past five years, this technology has seeped into politics, creating controversy about its ethics and place in elections.
1. The Argentine Presidential Candidate’s Obsession with AI
The 2023 Argentine Presidential election was called the “World’s First AI Election” by many because both candidates seemed obsessed with using AI to fight for the presidency. A fabricated picture, created using AI, portrayed Javier Milei’s opponent, Sergio Massa, dressed in an old-fashioned communist military uniform, his hand raised in a salute. Milei claimed to publish this photo to reveal how Massa’s campaign team used AI technology to capture the attention of citizens and sway their votes.
On an Instagram account named “AI for the Homeland,” Massa’s team posted several stylized AI-generated videos and photos depicting Meilei and his colleagues as barbaric zombies and pirates, attempting to defame his character. Pictures on the site showcased political posters with Massa as the hero, surrounded by crowds of smiling civilians. Darrell West, a senior fellow at the Center for Technology Innovation, told Connect News, “There were troubling signs of AI use in the election. Campaigners used AI to deliver deceptive messages to voters, which is a risk for any election.”
At the end of the battle between candidates, Melei won the runoff vote with 56% to Massa’s 45%.
2. Jason Palmer Utilized AI to Beat Joe Biden in the Samoa Democratic Caucus
Lesser-known politician, Jason Palmer, was a candidate in the 2024 Democratic Party Presidential Primaries and won the American Samoa Democratic Caucus against Joe Biden. The Presidential Primaries and the Democratic National Convention, held in March, are a public vote in which a presidential candidate for each party is nominated.
During the American Samoa Democratic caucus, Biden unexpectedly lost to Palmer, who allocated 51 votes to Biden’s 40. Palmer openly admitted to using generative AI during his campaign in the form of text bots that communicated with citizens through text or email. These bots allowed questions
about the presidential prospect’s policies and background. Along with bots, Palmer’s website featured an avatar named PalmerAI that mirrored his voice and identity to answer constituents’ inquiries.
“I believe I won partly because I’m very adept at technology,” Palmer explained to the Wall Street Journal. “I do think the people feel like I’ve been there because I did these virtual meetings, and I engaged with them using AI. I’m in favor of using technology to improve the world. And we shouldn’t stop using technology just because 10% of people are using it for bad things.”
3. India’s Prime Minister’s Campaign Speeches Translated in Real-Time Using AI
In 2022, India’s Prime Minister Narendra Modi took advantage of a new AI-powered tool, the National Language Translation Mission, created to translate Modi’s Lok Sabha campaign speech in real time, ensuring that his nationwide audiences could understand him.
Historically, language barriers have posed problems for Northern politicians trying to communicate with citizens from the South because of the multitude of languages spoken throughout Indian regions. The app, called “Bhashini,” meaning “speaker,” has become an innovative solution for
the Bharatiya Janata political party to communicate with its voters.
“Today, new technology has been used here through AI,” Modi said to the Times of India. “This is a new beginning, and hopefully,
it makes it easier for me to reach you.”
Modi has used advanced technology during his campaigns before. In 2012, the soon-to-be prime minister used a 3D hologram of himself to deliver speeches during the election polls in Gujarat, a vote in which he triumphed 116 out of the total 182 seats. According to VICE, he has made about 1,000 holographic appearances at nearly 900 rallies, reaching 8 million voters. Modi’s infatuation with technology, which is at the forefront of his campaigns, has proven successful, considering he has secured the title of Prime Minister of India since May 2014.
4. Prabowo Subianto’s Cute AI Dancing Videos Won Him His Presidency
Indonesia’s new president is a controversial ex-special forces commander who is disliked by many because of allegations of abuse, human rights violations, abduction, and even murder during a former president’s 30-year dictatorship.
President-elect Prabowo Subianto’s checkered past was overshadowed by his creative campaign tactic, an artificially generated dancing cartoon version of himself. Many are calling him gemoy, which is slang for “cute” or “adorable,” and refer to Subianto as Indonesia’s “cuddly grandpa,” a term that the country’s younger population has popularized. The 72-year-old politician won over many millennials and Gen-Z voters by using these AI-generated videos, according to the BBC. Yoes C. Kenawas is a research fellow at Atma Jaya University in Indonesia and noticed what Subianto’s campaign has achieved.
“I rarely see Prabowo’s real picture anymore,” Kenawas told the BBC. “This new avatar … is all over Indonesia. That’s how they’re softening his image. And so far, it’s pretty successful.”
Because many of his supporters were not alive at the time of his crimes and were ignorant of his past, Subianto was able to rebrand himself using social media, quickly influencing young voters with his new, lovable character. His team spent over $144,000 on social media advertising, a portion of which went to his AI avatar created by Midjourney.
In addition to Subianto’s avatar, a three-minute deepfake video of former dictator Suharto gained over 4.7 million views on X (previously Twitter). Suharto, who died in 2008, was a corrupt leader who was responsible for what international rights organizations claim to be one of the most brutal periods in Indonesia’s history. The deepfake was created by Golkar, a political party supporting Subianto. The intention of Golkar’s AI-generated video was obvious: to sway voters’ opinions towards Subianto. While initially somewhat convincing, citizens could identify that the media was fake.
5. Trump’s 2024 Presidential Win was Riddled with AI-generated Content
The 2024 presidential election in the United States was filled with AI-generated content advocating President-elect Donald Trump’s success.
According to CNN, the election saw more than 500 instances of viral misinformation, including false celebrity endorsements. Deepfake photos, generated by artificial intelligence, depicted Taylor Swift as Uncle Sam with the text, “Taylor wants you to vote for Donald Trump.” Others show “fans” of the pop star sporting “Swifties for Trump” T-shirts and claiming that after being targeted by ISIS, Swift fans want a “strong leader in the White House.” Donald Trump shared a collage of these images on his Truth Social account with the caption, “I accept!”
An AI-generated X post, shared in August by Trump, has amassed over 82 million views. It showed his opponent, Kamala Harris, addressing a crowd of Soviet-like men under a large communist banner. Elon Musk shared a similar photo of her. The caption read, “Kamala vows to be a communist dictator on day one. Can you believe she wears that outfit!?” Musk soon received heavy backlash for the post, having his image artificially manipulated to portray him as a communist figure.
On another occasion MAGA supporters created and reposted several fake images targeting Black voters. These images pictured Black citizens smiling for a group photo with the now-president. One creator of AI photos like these, Mark Kaye, hosts a Florida-based conservative radio show. He posted the photo to his Facebook account with an attached article about Black voters who supported Trump. To most of his one million followers, the photo seemed authentic, pushing what the co-founder of Black Voters Matter would call a “strategic narrative.”
Kaye told the BBC, “I’m not claiming it is accurate. I’m not saying, ‘Hey, look, Donald Trump was at this party with all of these African American voters. Look how much they love him!’ If anybody’s voting one way or another because of one photo they see on a Facebook page, that’s a problem with that
person, not with the post itself.”
There is no way to determine if AI is overall beneficial or harmful to democracy, but most politicians agree that there should be some regulation. Recently, META has required advertisers on their platforms to disclose when AI was being used to “alter or create political, social, or election-related
advertisements.” DeepMedia, a company that develops tools to uncover fake media, reported that globally, nearly 500,000 deepfake videos and audio recordings were shared on social media in 2023. A few generative AI companies are applying boundaries to their software, limiting its usage in political affairs. For example, OpenAI, the creator of ChatGPT, prohibits its image generator, DALL-E, from creating images with public figures, blocking the request with the warning that “it may not follow our current policy.”
Before the 2024 U.S. presidential election, Midjourney blocked users from creating fake content about Joe Biden and Donald Trump.
