On March 8, for the first time in Colombia’s history, an artificial intelligence candidate appeared on ballot papers across the country.
Gaitana IA (AI) ran for the Indigenous seat in the Senate and the House of Representatives in the northern state of Sucre.
While Gaitana did not win a seat in either of the country’s legislative bodies, it has sparked debate about the role of AI in Colombian politics.
With the ballots counted, Gaitana won a total of under 3,000 votes – less than 2% of the total votes for the Indigenous seat – suggesting that many people remain skeptical of this new digital approach.
Many questions have emerged surrounding Gaitana, such as why the Registraduría—the Colombian entity in charge of validating and accepting candidates—permitted this unprecedented candidacy, or what the intentions were behind the AI.
“Many local media outlets talked about an AI going to Congress, but that is not the case; they are humans leading the project,” Gaitana’s co-founder, Natalia Aase, told The Bogotá Post.
“It is actually a consensus tool developed by our community members, between 14 and 25 years old, from the Senú community of Reparo Torrente, in Coveñas,” she explained.
Rather than planning for the AI to assume office, Gaitana was devised as a democratic experiment underpinned by real human candidates
Aase detailed how the platform was designed to work: Colombian citizens could subscribe through a link to virtually participate and propose various debates regarding topics such as healthcare, women’s rights, and more. These interactions would also feed the AI database.
Once an initiative reached a collective consensus, the people occupying the seats in Congress would “decide the direction of the proposed laws.”
The two humans represented by Gaitana were Carlos Redondo Rincón, a Mechatronics Engineer from the Senú community, who was running for Senate, and Luz Rincón, an Embera-Katio Indigenous sociologist, who was seeking a seat in the House of Representatives.
The co-founder of Gaitana also revealed that the team conducted deep research into global democratic models, such as the one in Norway, and compared them with their own community dynamics.
As the research advanced, the team found that their community in Senú had already established a model of social interaction that worked well, prompting them to launch a digital project modeled on their own practices.
This meant digitizing their traditional way of reaching a consensus; in the Senú community, men, women, and youth gather around tables to discuss specific topics, such as women’s health or local fishing.
“Gaitana IA is not a generative AI; it is a participatory AI. What does that mean? Well, it is not ChatGPT. Instead, it takes the information provided by the users and organizes it,” pointed out Aase. “Transparency and security are the most important things for us; that is why we use blockchain technology—a system of blocks—to power this platform.”
According to Aase, the project was born from a motivation to prevent corruption and explained that with ‘Gaitana AI’, the decisions are not made by a single person but must be approved by at least 100 people.
“You might be able to manipulate one individual, but you cannot manipulate a hundred if you don’t even know who they are,” she concluded.