Is Argentina the first AI election?
The posters A certain Soviet aura spread across the streets of Buenos Aires.
There was one of Argentina’s presidential candidates, Sergio Massa, wearing a shirt with what looked like military medals, pointing to a blue sky. He was surrounded by hundreds of older people – in drab clothes, with serious and often disfigured faces – who looked at him hopefully.
The style wasn’t a mistake. The illustrator had been given clear instructions.
“Soviet political propaganda poster illustration by Gustav Klutsis showing a leader, masssa, standing firm,” was an indication that Mr. Mass in an artificial intelligence program was incorporated to produce the image. “Symbols of unity and power fill the environment,” the prompt continued. “The statue radiates authority and determination.”
Javier Milei, the other candidate in Sunday’s second election, has hit back by sharing what appears to be AI images depicting Mr Massa as a Chinese communist leader and himself as one cuddly cartoon lion. They have been viewed more than 30 million times.
Argentina’s elections have quickly become a testing ground for AI in campaigns, with the two candidates and their supporters using the technology to spoof existing images and videos and create others from scratch.
AI has made candidates say things they didn’t say and put them in famous movies and memes. Campaign posters have been created and debates have arisen over whether real videos are real.
The prominent role of AI in the Argentine campaign and the political debate it has sparked underline the growing prevalence of the technology and show that AI, with its growing power and falling costs, is now likely to be a factor in many democratic elections around the world.
Experts liken the moment to the early days of social media, a technology that offers tantalizing new tools for politics – and unforeseen threats.
Mr. Massa’s campaign has created an AI system that can create images and videos of many of the election’s key players — the candidates, running mates, political allies — doing a wide variety of things.
The campaign has used AI to portray Mr Massa, Argentina’s staid centre-left economy minister, as strong, fearless and charismatic, including videos showing him as a soldier at wara Ghostbuster And Indiana Jonesas well as posters evoking Barack Obama’s 2008 Poster ‘Hope’ and a cover of The New Yorker.
The campaign has also used the system to portray its opponent, Mr. Milei — a far-right libertarian economist and television personality known for his outbursts — as unstable, appearing in films such as “Clockwork Orange” And “Fear and loathing in Las Vegas.”
Much of the content is clearly fake. But a few creations have followed the line of disinformation. The Mass campaign produced one ‘deepfake’ video in which Mr. Milei explains how a market for human organs would worksomething he said philosophically fits his libertarian views.
“Imagine having children and thinking that each child is a long-term investment. Not in the traditional sense of the word, but thinking about the economic potential of their organs,” says the manipulated image of Mr. Milei in the made-up videoposted by the Mass Campaign on its Instagram account for AI content, called “AI for the Homeland.”
The caption of the post reads: “We asked an AI to help Javier explain the business of selling organs and this happened.”
In an interview, Mr. Massa said he was shocked when he first saw what AI could do. “I didn’t prepare my mind for the world I’m going to live in,” he said. “It is a huge challenge. We are on a horse that we need to ride, but we still don’t know the tricks.”
The New York Times then showed him the deepfake his campaign had created of Mr. Milei and human organs. He seemed disturbed. “I don’t agree with that usage,” he said.
His spokesperson later insisted that the post was a joke and was clearly labeled “AI-generated.” His campaign said in a statement that the use of AI is intended to entertain and make political points, not to mislead.
Researchers have long been concerned about the impact of AI on elections. The technology can mislead and confuse voters, creating doubt about what is real, contributing to the misinformation that can be spread through social networks.
For years, those fears had been largely speculative, because the technology to produce such counterfeits was too complicated, expensive and unsophisticated.
“Now we’ve seen this absolute explosion of incredibly accessible and increasingly powerful democratized toolsets, and that calculus has radically changed,” said Henry Ajder, a British-based expert who has advised governments on AI-generated content.
This year, a mayoral candidate in Toronto used bleak, AI-generated images of homeless people to telegraph what Toronto would become if he wasn’t elected. In the United States, the Republican Party posted a video made with AI showing China invading Taiwan and other dystopian scenes to depict what it said would happen if President Biden wins a second term.
And Florida Gov. Ron DeSantis’ campaign shared a video with AI-generated footage of Donald J. Trump attacking Dr. Anthony S. Fauci, who has become an enemy of the American right for his role leading the country’s pandemic response.
So far, the AI-generated content shared by the campaigns in Argentina has been labeled as AI-generated, or is so obviously fabricated that it is unlikely to mislead even the most gullible voters. Instead, technology has increased the ability to create viral content that previously would have taken teams of graphic designers days or weeks to create.
Meta, the company that owns Facebook and Instagram, said this week that it would require political ads to disclose whether they have used AI. Other unpaid posts on the sites that use AI, even if they are related to politics, would not have to make disclosures. So is the Federal Election Commission considering whether the use of AI in political advertising should be regulated.
The Institute for Strategic Dialogue, a London-based research group that studies internet platforms, signed a letter calling for such regulations. Isabelle Frances-Wright, the group’s head of technology and society, said the extensive use of AI in Argentina’s elections was worrying.
“I definitely think it’s a slippery slope,” she said. “A year from now, what already seems very realistic will only seem more realistic.”
The Massa campaign said it has decided to use AI in an effort to show that Peronism, the 78-year-old political movement behind Massa, can appeal to young voters by blending Massa’s image with pop and meme culture.
To do this, campaign engineers and artists put photos of Argentina’s various political players into an open-source software called Stable Diffusion to train their own AI system so it could create fake images of those real people. They can now quickly snap an image or video of more than a dozen top political players in Argentina doing almost anything they ask.
During the campaign, Mr. Masses of artists working with the campaign’s AI informed what messages or emotions they want to convey through the images, such as national unity, family values and fear. The artists then brainstormed ideas to incorporate Mr. Massa or Mr. Milei, as well as other political figures, into content that references films, memes, artistic styles or moments from history.
For Halloween, the Mass Campaign instructed its AI to create a series of cartoon-like images Mr. Milei and his allies as zombies. The campaign also used AI to create a dramatic movie trailerwith Buenos Aires, the capital of Argentina, burning, Mr. Milei as an evil villain in a straitjacket and Mr. Massa as the hero who will save the country.
The AI images have also appeared in the real world. The Soviet posters were among dozens of designs that Massa’s campaign and supporters printed to place in Argentina’s public spaces.
Some images were generated by the campaign’s AI, while others were created by supporters using AI, including one of the most famous: an image of Mr Massa on a horse in the style of José de San Martín, an Argentine independence hero.
“Mass was too stiff,” said Octavio Tome, a community organizer who helped create the image. “We show a boss like Massa, and he is very Argentinian.”
The rise of AI in Argentina’s elections has also made some voters question what is real. After a video of Massa looking exhausted after a campaign event circulated last week, his critics accused him of using drugs. His supporters quickly hit back, claiming the video was actually a deepfake.
However, his campaign confirmed that the video was indeed real.
Mr Massa said people are already using AI to cover up past mistakes or scandals. “It’s very easy to hide behind artificial intelligence when something you said comes out, and you didn’t want that to happen,” Mr. Massa said in the interview.
Earlier in the race, Patricia Bullrich, a candidate who failed to qualify for the runoff, tried to explain away the leaked audio recordings of her economic adviser offering a woman a job in exchange for sex by saying the recordings were fabricated . “They can rig votes and alter videos,” she said.
Were the recordings real or fake? It is unclear.