TSE regulates AI and prohibits deepfakes in municipal elections in 2024

O TSE (Superior Electoral Court) approved the regulation of the use of AI during the next elections, which will take place in October this year. See more details about the consequences of this event below!

Understand the TSE’s decision

O Superior Electoral Court approved yesterday, 27th, the regulation of the use of artificial intelligence in municipal elections, scheduled for October this year, in which the electorate will choose new mayors and councilors. The measure determines that there will be an obligation to warn voters about the use of AI in the production of content displayed in advertisements, and if the determination is not followed, the candidate may have their candidacy registration revoked or, if elected, they may lose their mandate. .

The decision also mentions the deepfakes, absolutely prohibiting its use, aiming to curb the production of fraudulent material using the image and voice of political candidates during the electoral campaign. In addition to deepfake, TSE mentioned the restriction of so-called chatbots, or virtual avatars, during contacts between campaigns and people, both by telephone and over the internet.

The idea of ​​the decision is to control and prevent the use of these technologies to create false recordings involving candidates and other authority figures during this year's electoral campaign, where their use could cause moral damage to victims of fraud and manipulation of the electorate's opinion. about them.  

Alexandre de Moraes, the current president of Superior Electoral Court, spoke about the new determination, saying that “all resolutions are important, but this is perhaps the most important to guarantee voters' freedom of choice and voting, and that in these elections, unfortunately, your excellency (mentioning minister Cármen Lúcia) will have to combat fake news and digital militias steroided by artificial intelligence”.

As for how the new determination will be reinforced, Moraes stated that the TSE will feature the most modern tools available today in order to combat the misuse of artificial intelligence and the spread of misleading messages.

“The TSE will have effective instruments to combat distortions in electoral advertisements, hateful, fascist, anti-democratic speeches and the use of AI to put something into a person’s speech that they did not say.” 

Alexandre de Moraes, president of the Superior Electoral Court

In conjunction with the new decision, the Court determined that social networks such as Facebook, X (formerly Twitter), Tiktok, among others, apply preventive measures to avoid the spread of publications with dubious content, containing unproven or out-of-context facts, with the obvious intention of deceiving the public who use these platforms. If this determination is not respected and posts containing hate speech and anti-democratic content, the networks will be held responsible and punishments will be imposed by the TSE.

How deepfake works  

As we have already seen in cases such as the propagation of videos with Explicit content involving pop diva Taylor Swift, on billionaire Elon Musk's platform, X, and the recent million-dollar scam suffered by a financial company in Hong Kong, deepfake has destructive potential in the hands of malicious people.

But to understand this power, it's worth having a quick explanation of how the technology works. Basically, it uses samples of a person's image and voice, analyzed by an AI, generating a database of the same. From that, the deepfake is capable of replacing one person's features with another, in real time.

Despite the common use of humorous satires such as those made by Jimmy Fallon on his American television talk show, where the comedian imitated the former president of the United States, Donald Trump, the deepfake can be used to apply scams and disinformation.

If the number of samples is sufficient, in addition to the image, it is also possible to replicate the voice of the identity that the person filmed is replacing in the new video. With the improvement of technology occurring at a dizzying pace, the potential of this AI-managed tool, both positive and negative, is increasingly greater, as the result becomes more realistic and even more difficult to be tracked and identified by victims in potential for digital crimes.

Voters must pay attention to the content 

Given the difficulty of identifying these cases of identity fraud, it is up to us, voters participating in this year's election, to take due care when consuming the content produced during the campaign, which will soon come into force throughout the country. 

It will not be enough to rely on the measures adopted by the TSE in curbing the production of material with the intention of deceiving those who will, in October, vote for the candidates who will represent us in city councils or city halls. There will be little care in the minefield that the broadcasting of commercials by candidates of dubious nature will become.

Despite this, the decision of the TSE is still important, as it shows that the government is aware and recognizes the vast destructive potential that the deepfake and technologies based on artificial intelligence have, taking this new determination into account. 

Sources: Capital letter, CNN Brazil e InfoMoney. 

LEARN MORE

reviewed by Glaucon Vital in 28 / 2 / 24.