Looking ahead to general elections of 2027, the Supreme Electoral Court (TSE) has begun to discuss for the first time how to confront one of the most recent challenges to democratic processes: the use of artificial intelligence (AI) and its potential to generate misinformation.
Although the proposal of create an observatory with support from universities has been well received, analysts warn that The biggest challenge will be to build clear rules for its regulation and have the institutional capacity to enforce them.
For Jahir Dabroy, analyst at the Research and Social Studies Association (Asies)the TSE does not start from scratch, since it has the Media and Opinion Studies Unit, created precisely to monitor this type of content, in addition to other agencies that have accumulated experience in the latest electoral processes.
“It is assumed that with two electoral processes they must have installed capacity, both in human resources and technical equipment.“he noted.
However, it considers that technical capacity will not be sufficient if the Court does not precisely define the scope of the measures it intends to adopt. In his opinion, The first step is to establish what the TSE will understand by disinformation.
“The key to all this is what they are ultimately going to call disinformation.“he stated. “If the same thing is not going to happen about what is an early campaign and what is not an early campaign… it seems that it depends on who I play.”
In this sense, Dabroy maintains that the Court will have to decide whether any content created with artificial intelligence will be subject to restrictions or only that which is intended to deceive the electorate. He recalled that AI can also be used to produce legitimate content, so a blanket ban could generate new controversies.
“I can ask valid questions with artificial intelligence“he explained.
Another aspect that, in his opinion, must be clearly defined is who will be subject to those rules. The analyst considers it essential to determine whether the restrictions will fall only on the political parties, on the candidates or on both, to avoid gaps that later give rise to different interpretations.
“Establish whether it goes to the parties or the candidates or goes to both as well. These are issues that seem minimal, but if they are not well defined, they give rise to various interpretations and that delegitimizes the work of the Supreme Electoral Tribunal.“he indicated.
Likewise, it considers that these provisions should be formalized through a regulation or an agreement of the TSE plenary session to provide legal certainty regarding their application.
In addition to regulation, Dabroy identifies another challenge: the ability to trace the origin of content manipulated through artificial intelligence, such as so-called deepfakes. It warns that a fake video could be created by third parties with the purpose of harming a political party, making it difficult to determine who is really responsible for its dissemination.
Added to this is the speed with which this type of material can go viral on digital platforms, which is why an immediate response from the authorities is considered essential.
“A video went really viral in a matter of a couple of hours… it does have to have a responsiveness“, he noted, adding that the TSE must coordinate closely with the Electoral Crimes Prosecutor’s Office of the Public Ministry to investigate the origin of the content and act in a timely manner.
Renzo Rosal agrees that the fact that the Court incorporates artificial intelligence as a risk factor for the first time represents progress, although he warns that the challenge far exceeds the current capabilities of the institution.
“It seems to me that the fact that the Court touches on the risk factor linked to artificial intelligence for the first time is good.“he stated.
However, he considers that the speed with which this technology evolves will make it very difficult for the electoral body to respond with the same speed. Added to this is the lack of specialized experience in the country to technically support the announced actions.
“I do not see that any university in Guatemala is specifically working on areas of work linked to artificial intelligence.“he commented.
Rosal maintains that monitoring should fall mainly to the Specialized Unit for Media and Public Opinion Studies and the IT Unit of the TSE. However, he warns that the main obstacle is the lack of specific regulation within the Electoral and Political Parties Law.
“The electoral law obviously does not touch anything regarding artificial intelligence… so the parties have an open field“, he expressed.
The analyst considers that the main risk does not fall solely on the electoral authority, but on the citizens themselves, who could have difficulties distinguishing between authentic content and material manipulated by artificial intelligence.
“The main impact for me is not on the side of the Court, but on the side of the citizen“he stated.
In his opinion, the institution will face difficulties in keeping up with technological development.
“The Supreme Electoral Court is going to be permanently overwhelmed by this reality“, he assured. “Technological dynamics are moving at an enormous speed and the Court has been characterized as a very conservative institution.”
Rosal adds that this situation is similar to the management of the early campaign, where, although the TSE has legal tools to impose sanctions, it considers that it has not applied them firmly.
“The number of files does not matter; What matters is that the Court exercises authority and imposes sanctions“he concluded.
For both analysts, the real challenge lies not only in creating an observatory or incorporating new technological tools, but in establishing clear rules on the use of artificial intelligence, providing them with legal support and guaranteeing that the TSE has the capacity to apply them quickly and with certainty during the 2027 electoral process.