On November 1, 2026, O Globo, the Rio de Janeiro daily, reported that Luiz Inácio Lula da Silva’s government is racing to get Congress to pass two bills (PL, in the Brazilian acronym) before its composition changes. The first is the legal framework for artificial intelligence, PL 2338/2023, which the Senate has already approved and which is awaiting a report in the Chamber of Deputies. The second is PL 4675/2025, which gives CADE (the Administrative Council for Economic Defense, Brazil’s competition authority) the power to impose rules in advance on the largest digital platforms. That one is already ready for a floor vote.
If both pass, AI regulation in Brazil will combine something that Argentina and Chile lack: a risk-tiered AI law, with compensation for copyright holders, and competition obligations set in advance for big tech. The government presents both bills as matters of competition and digital sovereignty, and leaves content moderation out of the picture. According to O Globo, the next Congress will have “greater participation from lawmakers aligned with the right.”
Brazil’s AI law (PL 2338/2023)
PL 2338/2023, authored by Senator Rodrigo Pacheco, was approved by the Senate on December 10, 2024, and reached the Chamber of Deputies on March 17, 2025. It classifies AI systems by their level of risk, refers liability for damages to the Civil Code and the Consumer Protection Code, and requires anyone training systems on copyrighted works to compensate rights holders. The approved text sets fines capped at R$50 million per violation; for private companies, the fine is calculated at up to 2% of the group’s gross revenue in Brazil, without exceeding that cap (Art. 50, II). The text also provides for suspending the system and banning the processing of certain databases.
The bill is listed as “Awaiting Rapporteur’s Report in the Special Committee”: the rapporteur, Deputy Aguinaldo Ribeiro (PP-PB), has not yet submitted his report. On August 24, Ribeiro told Valor Econômico, the Globo group’s business daily: “We’re in a hurry, because I think this window of opportunity will close fast.”
The big tech competition bill (PL 4675/2025)
The Executive Branch submitted PL 4675/2025 on September 18, 2025. It creates a superintendency within CADE that can designate “economic agents of systemic relevance in digital markets.” The first threshold is annual gross revenue: more than R$50 billion worldwide or more than R$5 billion in Brazil. Meeting either one is enough. O Globo wrote that both were required, but the original text and the rapporteur’s substitute say “or.” Crossing the threshold only triggers the review: CADE then evaluates qualitative criteria, such as third-party dependency or control over data. And the designation does not bring automatic obligations. O Globo gives an example of a possible obligation: that an operating system not favor a single digital wallet.
The Chamber approved fast-track status on March 18, 2026. On July 8, the rapporteur, Deputy Aliel Machado (PV-PR), submitted a substitute bill that limits the designation to six years, down from ten in the original. Machado told O Globo: “Competition rules apply to other sectors, and they have to apply to this one too.” According to the newspaper, the Finance Ministry and CADE viewed the changes favorably, and they were expected to be discussed at this week’s leadership meeting.
Why now
In his victory speech, quoted by O Globo, Lula said: “We had to overcome outside interference coming in, through digital platforms that we’re going to have to regulate” (our translation). According to the newspaper, Meta, owner of Facebook, Instagram and WhatsApp, acknowledged technical glitches that temporarily affected accounts belonging to the campaign of the PT, Lula’s party, and denies having interfered to favor Flávio Bolsonaro, son of former president Jair Bolsonaro. In a statement on October 16, Meta says the Lula campaign’s ad account was deactivated “by mistake for just over an hour” and that it “does not take sides in this or any other election” (our translation). Meta, Google and Amazon did not respond to O Globo.
How Brazil compares with Argentina and Chile
Argentina has no AI law. On October 20, Federico Sturzenegger, the deregulation minister, said according to La Nación: “We’ve decided not to regulate preemptively. We’ll regulate if a problem arises.” The official strategy is going a different route, as we explained in our analysis of automated companies.
In Chile, the AI bill (bulletins 15.869-19 and 16.821-19, merged; the bulletin is the number the Chilean Congress uses to identify each bill) is in its second reading in the Senate. According to the Senate’s records, the Future Challenges Committee approved it in general terms on October 26 and sent it to the floor. According to Fundación Progresa, the government announced a substitute amendment that it has not yet submitted.
What to watch
First, whether the leadership meeting sets a date to vote on PL 4675/2025. Second, Ribeiro’s report: without it, PL 2338/2023 cannot leave committee, and after that it still faces a floor vote in the Chamber of Deputies and, if amended, another pass through the Senate before the new Congress is seated.
Frequently asked questions
Does Brazil already have an artificial intelligence law?
Not yet. Brazil’s AI legal framework, PL 2338/2023, was approved by the Senate on December 10, 2024, and has been in the Chamber of Deputies since March 2025, sitting in a special committee awaiting the report from rapporteur Aguinaldo Ribeiro. After that, it still needs a floor vote, and if the Chamber amends it, it would go back to the Senate.
What fines does Brazil’s AI law set?
PL 2338/2023, as approved by the Senate, sets fines capped at R$50 million per violation. For private companies, the fine can reach up to 2% of the gross revenue of the company, its group or conglomerate in Brazil in the last fiscal year, before taxes, always within the R$50 million cap per violation. The text also provides for suspending the system and banning the processing of certain databases.
Which companies would Brazil’s big tech regulation cover?
PL 4675/2025 allows CADE, Brazil’s competition authority, to designate as “of systemic relevance” groups with annual gross revenue of more than R$50 billion worldwide or more than R$5 billion in Brazil. Meeting one of the two thresholds only triggers the review: CADE also evaluates qualitative criteria, such as third-party dependency or control over data.
What’s the latest development on AI regulation in Brazil?
According to O Globo, as of November 1, 2026, Lula’s government was trying to pass both the AI law (PL 2338/2023) and the big tech competition bill (PL 4675/2025) before Congress’s composition changes. The second one is ready for a floor vote. The AI law is awaiting the rapporteur’s report in a special committee of the Chamber of Deputies.
Does Brazil’s AI regulation include content moderation?
No. According to O Globo, neither the AI law (PL 2338/2023) nor the big tech competition bill (PL 4675/2025) regulates content, and Lula’s government deliberately decided to leave that issue out. Both bills are presented as rules on risk, copyright and competition, with CADE as the authority over the largest digital platforms.