Is the use of artificial intelligence in elections a problem? – 02/01/2024 – Education

If the 2022 Brazilian elections represented a watershed in the dissemination of misinformation and, consequently, in the fight against it, this year’s election brings a new and even more unpredictable challenge: the use of artificial intelligence (AI) tools.

The discussion around this technology and its impact has been broad in all sectors of society, but it becomes even more important in times of elections. It was with this aim that the Superior Electoral Court (TSE), led by minister Cármen Lúcia, rapporteur of the rules for electoral advertising this year, held a Cycle of Hearings in the second half of January that included almost a thousand suggestions.

On the third and final day of the cycle, one of the topics that generated the most interest was precisely the one dealing with the rules for the use of artificial intelligence in elections.

Allow or prohibit, identify or not, regulate or leave free? All these questions were raised by participants from the most diverse sectors, including technology experts, representatives of civil society, political parties and advertisers responsible for electoral campaigns.

The concern about the use of deepfakes —synthetic media created through the manipulation of images and sounds— and other types of fraudulent messages that aim to confuse the electorate is real and strongly justifiable, as examples of improper use of this type of media are already beginning to emerge. tool.

One of the most recent cases involves the pre-candidate for mayor of Maringá, in the interior of Paraná, who was the victim of false content that used a voice similar to his, but declaring support for an opponent and indicating his withdrawal from running.

Video and social media platforms already prohibit the broadcasting of technically manipulated content that aims to deceive users, including in advertisements. But the question is how to ensure that this practice is not used to circulate videos outside channels with pre-established rules. How to check a credible video that arrives via instant messaging service, sent by people close to you, without an identified origin or source?

It is in this scenario that one of the possible solutions is precisely the use of artificial intelligence tools to detect manipulated content. As the representative of the Brazilian Institute for the Regulation of Artificial Intelligence, Marcelo Senise, put it, “the TSE must invest in deepfake detection technologies and AI algorithms that identify suspicious patterns of dissemination of disinformation.”

Another path, as important as regulation for the use of these technologies, is to invest in public awareness and citizen education so that we learn to be suspicious of this content and identify its origin, source, context and purpose. Having a critical sense is essential as images become increasingly difficult to check for veracity.

The development of these skills is the reason for media education, which must, at crucial moments such as elections, be offered in conjunction with education content for democracy. It is with this objective that the Instituto Palavra Aberta launches the new version of the program #FakeTôFora: Quem Vota se Informa, with content aimed at high school students and which offers a learning path adaptable to different subjects and teaching stages.

The sustainability of our democracy also depends on investing in those who will strengthen and defend it. Artificial intelligence poses a challenge that will become increasingly complex in the short and medium term. Without educating ourselves about information, valuing ethics and responsibility, this path can be even more tortuous.