In a video circulating on social networks in Brazil you can see the famous singer Anitta advertising a gambling game: “I won 300 reais the other day“, states the voice of the artist. In another spot, Neymar, the soccer star, also advertises the platform. But both videos have something in common: They are montages made with Artificial Intelligence.
The voices and images of both figures were manipulated with tools of Artificial Intelligence (AI) and this aroused the concern of the Brazilian authorities to find urgent answers, especially in the face of October municipal elections. Already in February, the Superior Electoral Court (TSE) approved a proposal that establishes rules for the use of AI in the electoral context.
Brazil and the problems with Artificial Intelligence
Both Neymar’s and Anitta’s team of advisors accused that the videos were “fake news“and that the player and the artist They were never part of said advertising. The creation of fake photographs, videos and voice audios of celebrities It is an increasingly bigger problem, which in recent times has even led to Taylor Swift to a dispute with the social network “X”.
In the face of this problem, the panorama in Brazil is complex. The Latin American giant is one of the most connected countries in the world – with more cell phones than inhabitants – and in recent years he was a victim of strong disinformation campaignswhich even reached the political level.
An example of these “fake news” campaigns is what happened with the former president Jair Bolsonaro who in 2023 was declared ineligible for eight years for abuse of power and misuse of the media having stated, without evidence, that the Brazilian electoral system was not secure. With the exponential development of AI technologies, disinformation campaigns They are becoming stronger and indistinguishable and put at risk the democratic coexistence of the countries.
“Fake videos can be used to manipulate public opinion, defame individuals, and interfere with democratic processes.“said the Brazilian IT Ana Carolina da Hora. Furthermore, last Wednesday the NGO Center for Countering Digital Hate warned about the risks of AI in a year marked by electoral contests in the world, upon seeing the ease with which they can create fake images of the candidates in the United States elections.
In Brazil, electoral defamation is already a reality. During 2023, the police identified a case of assembly against the mayor of the city of Manausin which a audio with the manipulated voice of the politician who intended to insult teachers in the municipality.
Brazil: first steps in AI regulation
With the municipal elections on the horizon, the discussion about the use of the Artificial Intelligence mobilizes authorities and legislators in Brazil. Already in February, the Superior Electoral Court (TSE) – who disqualified Bolsonaro and whose mission is to guarantee transparent and democratic elections -, approved a proposal that establishes rules for the use of AI in the electoral context.
INTER – LATAM-POLITICS_opt.jpeg
The rule will be applied in the next elections and stipulates: ban the use of “deepfakes” (technology that allows faces and voices to be manipulated) and restrict the use of “chatbots” and avatars in the campaign.
In addition, it establishes that the manipulated content must be identified. A decision similar to the one made by the company GOAL and that it will seek to implement in the coming months.
Alexandre de Moraes, president of the TSEnoted that the court’s challenge will be “combat fake news and digital militias anabolicized by artificial intelligence“In addition, de Moraes assured that the use of technology and montages can serve to “change the election result“.
Also, the Congress of the country also got into the debate, but with the focus on regulating the different applications of AI in the social and economic life of the country. Thus, one of the proposals under discussion – raised by the president of the senate Rodrigo Pacheco – provides for the regulation of the different scenarios in which AI can be applied, although it remains to be seen how this regulation will be carried out.
Source: Ambito
