Colombia is experiencing its first presidential campaign, with each candidate having artificial intelligence at their fingertips, but the ways each candidate uses it is markedly different. While almost all of them focused on AI-generated videos, social media was flooded with the results of two campaigns above all else.
First, the far-right candidate, criminal lawyer Abelardo de la Espriela, is known as “El Tigre” (The Tiger). He dances on stage in front of a video of a big cat and transforms into a giant tiger that destroys trains that are supposed to represent traditional politics.
Sen. Paloma Valencia, a traditional right-wing candidate, is also bringing to the campaign several “fruit novels” (a viral phenomenon on TikTok) featuring characters such as apples, bananas, and chili peppers. In frutimbiathe senator is a cheerful pineapple that stands out against the banana and de la Espriela coconut on the left.
AI-generated videos are cheaper to produce than traditional campaign ads, meaning they can be used by all candidates, not just the frontrunners. People who are further behind in the polls are using them as well, with centrist candidate Sergio Fajardo creating an imitation of a character from the movie. harry potter. Mauricio Liscano, a former minister under Colombian President Gustavo Petro, created a version of the film. encanto of the title Descencant. “They live in a surreal Colombia,” the video says of the rivals.
Even leftist Sen. Iván Cepeda, who is leading the polls in a more traditional campaign, released an AI video last week in which he says his refrigerator is “happy because it’s fully stocked,” a nod to the government the candidate supports.
Most of the AI videos share a decidedly artificial aesthetic, such as dancing tigers, fruit, or Disney-like characters. They do not try to pass it off as truth or mislead. They are clearly caricatures.
“These videos are not trying to show reality, they are trying to show emotions,” says Juan Diego Castañeda, co-director of the Charisma Foundation, an NGO that monitors the use of technology by the powerful. “Politics has long valued emotion over proposition, and these videos follow the same visual logic we’ve seen for decades, from movies to ads. The difference is that candidates are now saving a lot of money thanks to AI.”
For example, Mr. de la Espriera’s dozens of videos, including one showing a tiger walking through a crowd, are clearly fake. But that doesn’t stop them from appealing to voters’ emotions, especially with a narrative in which lawyers are the only candidates outside the political establishment.
These videos were so effective that members of the traditional right complained about the depiction of former president Alvaro Uribe in these videos. In it, he is depicted as allying with political opponents such as former Bogotá mayor and presidential candidate Claudia López, former president Juan Manuel Santos, and Interior Minister Armando Benedetti. It’s a fantastical image and a conspiracy theory that can’t be imagined outside of AI. But it still creates a narrative that divides the world into “them” and “us.”
In the campaign world Valencia has created, she presents herself as someone who can bring disparate groups together, like a fruit salad. Petro’s supporters are rotten bananas, and Abelardo de la Espriela is a plantain disguised as a coconut. The idea that Petro and de la Espriela supporters are on the same team also has no basis outside of AI.
In another video, her rivals appear like ears of corn from which popcorn has popped out after being struck by vice presidential candidate Juan Daniel Oviedo’s “lightning bolt.” These are the center-right voters the senator is aiming to advance to the second round.
AI videos have reinforced stereotypes about the election, such as portraying Sergio Fajardo as an indecisive person and Abelardo de la Espriela as a robot who can only repeat “loyalty to the fatherland.” But they also reproduce a far more problematic trope in Colombian history: a depiction that would be unacceptable outside of caricature.
For example, in some videos by Valencia and de la Espriela, Cepeda is depicted as a guerrilla fighter or guerrilla ally. “I think this is worrying because it reinforces prejudice and persecution against left-wing parties that are accused of collaborating with guerrillas,” says the charismatic Castañeda.
The exception to the trend toward AI seems to be Cepeda, the candidate who is leading in the polls. Other than his recent talking fridge, he hasn’t posted videos of fruit or animals on social media, instead focusing on public square speeches. “I’m not surprised at all. His campaign probably believes that voters aren’t online or using social media,” says Carolina Botero, an independent consultant on artificial intelligence and a member of the charismatic board of directors.
“The contrast is very reminiscent of the 2010 green wave,” she added. This is a reference to former Bogotá mayor Antanas Mocs’ presidential campaign, which focused on social media messaging. Mockus made it to the second round, but lost to Juan Manuel Santos, who was boosted by the popularity of Álvaro Uribe (while they were still in an alliance) and spent more time traveling domestically.
“We don’t know if that strategy will work for Mr. Cepeda, but on the other hand we know that Mr. de la Espriela’s voters are very active on social media, where he and Mr. Paloma are having heated debates and creating a lot of buzz. We need to show who is generating more movement,” she says.
Analysts interviewed agreed on one piece of good news: While these videos are emotionally charged, they have not turned into the massive disinformation machine that many feared, at least not yet. “Our constant fear of surreal videos tricking us has not resulted in a major scandal, at least so far,” Botero says.
There are exceptions. Since March 8, the Columbia Check website has identified an AI-generated video showing Petro voting in the party’s internal primaries despite calls for him to abstain. But overall, AI videos in Colombian elections still struggle to convincingly imitate candidates.
“Although AI-generated video has improved significantly, it still has a long way to go before it becomes a reality,” says journalist Juanita Vélez, co-director of MioPÍA, a platform that monitors the use of this technology in Colombia. “It’s much easier to fool people using voice,” adds my colleague Daniela Amaya.
As an example, they point to an audio clip shared by the Minister of Labor that purports to show Senate President and opposition politician Efrain Cepeda plotting fraud with the National Registrar ahead of parliamentary elections. A similar audio clip released days before Slovakia’s presidential vote resulted in the liberal candidate losing the race. In Colombia, AI-generated audio and video have never had such a decisive impact (so far). As with past campaigns, there were memes and billboards that spread falsehoods.
So far, AI has shaped campaigns primarily through videos that sell fantasies and emotions. It is important to spread this battle, but spreading does not guarantee victory. Former Petro minister Mauricio Liscano, who has been trailing in opinion polls, also appears as an adorable cat in one such video. “Kittens have up to 300% more influence than hate,” says the animal, promoting his proposal. The clip has been viewed 2.7 million times on YouTube.
register to our weekly newsletter Get more news coverage in English from EL PAÍS USA Edition
