Disinformation entered the Ecuadorian presidential election with a new element: the advanced use of artificial intelligence (AI) to create deceptive candidates.
Altered videos with remarkable realism are the result of an advanced, costly, and exceptional approach, according to experts, surpassing strategies in neighboring countries.
The AI has been utilized in manipulated content for the upcoming presidential election, mimicking the voice and lip movements of candidates or appearing as a television presenter to damage their reputation.
Leftist candidate Luisa González and president Daniel Noboa, who were popular in the polls, were repeatedly attacked in these messages that were circulated thousands of times, as confirmed by AFP’s digital verification team.
Days after Nicolás Maduro started his third term (2025-2031) in Venezuela, a video of González endorsing his inauguration went viral on social media, praising him as a “great president.”
When Donald Trump became the President of the United States, two videos circulated widely showing him alleging that Noboa and his family’s banana business had connections to drug trafficking.
Both statements are manipulated.
“The AI’s significance lies in its realistic quality, as stated by AFP Marco Benalcázar, who oversees the Alan Turing Research Laboratory in Artificial Intelligence and Vision at the National Polytechnic School of Quito.”
Technology enables the replication of a person’s voice and the synchronization of lip movements, a form of manipulation referred to as “deep fake.” The duration of the process, which can be completed within an hour, varies based on the complexity of the setup.
Users of the internet, who have become more skilled at recognizing edited photos, are now encountering incredibly lifelike videos “in a fast-paced environment where they lack the time or means to verify everything,” according to Carlos Rodríguez Pérez, a professor at the Faculty of Communication at the University of La Sabana in Colombia.
A carefully planned approach –
What motivates these campaigns and what is their objective?
This type of carefully crafted content necessitates strategic planning when choosing which politician to use as a voice and determining the subject matter. These decisions are deliberate and not arbitrary, according to Rodríguez.
The expert stated that the purpose of the content is to undermine the credibility of the opposition by associating candidates with criminal activities or contentious ideas, such as a doctored video implying that González supports state-funded gender transitions for children.
Implementing these strategies necessitates a significant financial investment.
“When you enlist the services of a marketing firm, you have the option to compensate individuals for creating inauthentic material with a specific goal in mind,” Benalcázar points out.
A fake message about González was spread on a website designed to mimic his party’s site. The manipulated video featured the candidate introducing a “card” for services like food, health, and education, resembling Venezuela’s “carnet de la patria.”
The production centers of “deep fakes,” referred to as “fazendas” by Rodríguez, are capable of impersonating journalists accurately. In one instance, a presenter in an alleged NTN24 report claims that Noboa paid one million dollars to attend Trump’s inauguration. Another instance shows Fernando del Rincón from CNN linking González to criminal activities.
Responsible utilization
Three bills presented in the National Assembly aim to restrict the scope of this technology while also promoting its responsible use due to concerns raised about its extent.
AI, which experts such as Rodríguez believe can effectively address various issues, should not be vilified.
“The use of technology determines whether it is good or bad,” according to Benalcázar. He also mentions that disinformation is one of the negative ways it can be used.
Candidates frequently mentioned the AI during the January 19 presidential debate as a way to execute their government programs, but they did not provide specific technical information.
During the debate, Christian Caldas, a 32-year-old designer who was getting lunch in northern Quito, was disappointed that no one presented a compelling idea and instead focused on criticizing each other, particularly about AI.
“The AI is incapable of running a nation; people are essential for governance,” he remarked, expressing his disbelief.