Lawmakers who sit on the House commission analyzing the AI (artificial intelligence) regulation bill in Brazil’s Congress acted in favor of big tech companies after receiving representatives of firms with interests in the AI sector in their offices.
This conclusion is based on an exclusive investigation by Aos Fatos, which shows that big tech lobbyists visited the House at least 83 times between the establishment of the commission discussing the bill 2.338/2023 on May 20, 2025, and October of last year.
Approved by the Senate in December 2024, the bill creates a regulatory framework for AI and includes provisions that run counter to big techs’ commercial interests, such as increased corporate liability for the technologies they put on the market.
The bill was expected to be voted on by the House at the end of last year, but the report was not released due to a lack of political momentum for a vote, according to reporting by Brazilian news outlet Jota. The chair of the commission discussing the matter, Representative Luísa Canziani (PSD–PR), says the text should be considered later this semester.
The visits appear in the House’ access records, obtained through Brazil’s Access to Information Law (LAI), a legal instrument that allows citizens to request information from public entities. Entries by employees working in government relations for Amazon, Google, IBM, Meta, Microsoft, and OpenAI were analyzed.
The records indicate that 15 offices of commission members received lobbyists during the period analyzed, in addition to leadership offices and hearing rooms used to discuss regulation. Once inside the House, however, visitors can circulate freely, making it impossible to determine the full extent of their contacts.
Lobbying is not illegal in Brazil, but the lack of regulation makes the activity opaque in the country — unlike in the United States, where lobbyists are required to disclose which issues they advocate for, who their clients are, how much they spend, and whom they meet with.
The delay in concluding lobbying regulation — pending in Congress for nearly two decades — undermines oversight of the influence of private interests on public policy debates, which often grows in proportion to the economic stakes involved.
“While we can only go once, paying out of our own pockets, they are there every week,” a civil society representative from Brazil who asked not to be identified told Aos Fatos. “That violates the principle of a level playing field,” they added.
Despite the limited transparency, analyzing the actions of lawmakers who received big tech representatives in their offices helps demonstrate political alignment with corporate interests. In some cases, lawmakers’ actions occurred on the very same day the visit was recorded.
Adriana Ventura (Novo–SP)
On September 30 of last year, federal representative Adriana Ventura (Novo–SP) — vice chair of the commission analyzing bill 2.338/2023 — took the floor during a committee hearing to argue for fewer restrictions on the use of artificial intelligence in public security, claiming that looser rules would benefit the population.
The rhetoric that AI can help fight crime has been adopted by OpenAI in its conversations with Brazilian authorities, but the coincidence does not end there: on the same day she spoke in Congress, Ventura received OpenAI’s chief economist, Aaron Chatterji, in her office.
The lawmaker had just returned from an official commission mission to the United States, during which she met with OpenAI executives in Silicon Valley, according to her own social media posts. The tour included visits to the headquarters of Anthropic, Nvidia, Palantir, Scale AI, Apple, xAI, and 1X Technologies, as well as Stanford University.
Upon returning from the trip — similar to one she took the previous year —, the lawmaker listed what she called the “immediate lessons” learned from conversations with big techs, concluding that AI regulation “cannot be a straitjacket” and must be “proportional, revisable, and protect rights, but also allow Brazil to compete and advance”.
In late October, during a public hearing on the use of AI in agriculture, Ventura again warned of the risk that regulation could “make creative solutions unviable,” echoing arguments that technology companies have used for more than a decade, as Aos Fatos has already shown.
An analysis of Congress access records also indicates that Ventura had already received a visit from an OpenAI lobbyist in July 2025 — the same day she introduced a request to invite experts to the commission’s public hearings.
That request included, for example, Professor Juliano Maranhão, of the Lawgorithm Institute, who testified before the AI Commission on September 2. In his remarks, the guest questioned whether AI training violates copyright and argued for lower development costs — positions aligned with big tech interests.
AI regulation is not the only issue on which the lawmaker has sided with technology companies. Ventura opposed a bill that would have criminalized electoral disinformation and worked to remove sections of legislation proposing measures against the sexualization of children and adolescents online.
Vitor Lippi (PSDB–SP)
Congressman Vitor Lippi (PSDB–SP) is the current president of the Parliamentary Front for the Machinery and Equipment Industry, a group of lawmakers that considers the adoption of artificial intelligence by industry one of its priorities.
Lippi received Nicolas Robinson Andrade, OpenAI’s head of public policy for Latin America, on September 2. On the same day, he argued before the AI commission that Brazilian regulation should be more lenient to protect AI’s potential for research and innovation.
“I know we always want to be innovative, but we need to be very careful, because these things migrate from one place to another, just like investments,” Lippi said.
The same rhetoric had already been used by Andrade. “Investments — whether in model training, opening operations of large and small companies, or infrastructure — will flow toward countries with favorable regulatory environments,” the OpenAI executive said in June.
At the September 2 session, the House commission discussed the relationship between generative AI and copyright. The issue is one of the most sensitive points for big tech companies, which have been accused by Brazilian artists, journalists, and other professionals of using their works to train language models without authorization or compensation.
At the hearing, Lippi’s position also reflected that of the companies. “We need to have common sense so as not to harm those in culture, but also not increase costs or uncertainty for those developing new solutions for humanity,” he said.
In a statement sent to Aos Fatos (see full text below), Lippi confirmed the meeting with the lobbyist but said it took place after the public hearing and “did not influence the congressman’s position on the issue”.
In addition to the AI commission, the lawmaker has also defended big tech interests in sessions of the Committee on Science, Technology, and Innovation, of which he is also a member.
Jadyel Alencar (Republicanos–PI)
Congressman Jadyel Alencar’s (Republicanos–PI) proximity to big tech companies dates back several years. In March 2025, he posted videos with Google and Meta employees announcing discussions with the companies about new partnerships for his home state of Piauí.
In 2024, the lawmaker had already organized a training course in metaverse and augmented reality with Mark Zuckerberg’s company and, last year, repeated the partnership to promote a digital entrepreneurship event in Piauí's capital Teresina.
On June 10 of last year, Alencar received three Google lobbyists in his office, according to House access records. At the time, he was rapporteur of bill 2.628/2022 — the lawmaker responsible for drafting the official report on the proposal — which provided protections for children and adolescents in digital environments and was enacted in September of last year.
In August, the bill became known as the “Adultization Bill,” following a complaint by Brazilian influencer Felca about the inappropriate exposure of minors on social networks.
Even under public pressure after Felca’s video went viral, the lawmaker accepted suggestions from Google and Meta in his report: he removed or altered sections addressing platforms’ “duty of care” or their obligation to actively prevent children and adolescents from accessing inappropriate content. These changes were approved and enacted.
In an interview with Brazilian news outlet GloboNews, Alencar argued that, in his view, children and adolescents should not use social networks, but said he sought a middle ground in the proposal “with the development sector and the economy in mind”.
After approval of the so-called Digital Child and Adolescent Statute, the lawmaker continued to defend big tech interests in discussions of Provisional Measure (MP) 1.318/2025, which creates Redata (Special Tax Regime for Data Center Services in Brazil) — MP is a type of executive order issued by the federal government that has the immediate force of law but requires subsequent approval by Congress to become permanent.
In September, Alencar introduced three amendments proposing a temporary import regime (see here, here and here) — essentially a tax exemption — for data centers, arguing that these structures require “constant technological updating.” In another amendment, he proposed using surplus energy to power such facilities.
In a statement sent to Aos Fatos, Alencar said he engages in dialogue with all stakeholders and that his amendments are not aligned with big tech interests.
Luísa Canziani (PSD–PR)
The chair of the AI commission has a history of closeness with major technology companies. She previously led the Digital Front, a parliamentary group linked to big techs, whose inaugural speech was delivered by Google Brazil president Fábio Coelho.
In April of last year, Aos Fatos revealed a video posted on her Instagram profile in which Canziani praised Google and referred to a Brazilian employee of the company as a “friend” during a visit to one of its U.S. headquarters. Google is also a partner of the Paraná State Secretariat for Artificial Intelligence, headed by former federal deputy Alex Canziani (PSD–PR), the congresswoman’s father.
IBM, another partner of the secretariat led by her father, sent an employee to Canziani’s office on June 16 of last year, according to Aos Fatos.
As expressed in a public hearing in September, the multinational has sought to loosen the risk classification approved by the Senate in the vote on bill 2.338/2023, which sets stricter rules for AI models that may impact fundamental rights.
Days after that hearing, Canziani also participated in a Free Market Parliamentary Front tour that took Brazilian lawmakers to the United States to discuss AI with big tech representatives. On social media, she said the trip was not funded by the House, but did not disclose who sponsored it.
That information contrasts with a statement by Rep. Aguinaldo Ribeiro (PP–PB), the bill’s rapporteur in the House, who declared the trip an official mission and published a report detailing the meetings held in the United States.
The document shows, for example, that OpenAI “presented its vision of AI aligned with a historical context, comparing it to electricity to argue that it represents a change of similar magnitude”.
One month after the U.S. mission, Canziani published an article in Jota with André Godoy, director of the development agency ABDE, comparing resistance to technology to the skepticism that arose during the Industrial Revolution and stating that “AI already occupies spaces once reserved exclusively for human cognition” and that, as in the past, “there is no way to stop this advance”.
Representatives’ statements
Aos Fatos contacted the offices of the lawmakers mentioned in the report and sent messages to their communications teams. Lippi, Alencar and Canziani responded with the statements reproduced above.
Statement by Vitor Lippi
The office of Federal Representative Vitor Lippi clarifies that the meeting mentioned did in fact take place on September 2 at 5:30 p.m., therefore after the public hearing of the AI Commission, and had an institutional and transparent character, like many other meetings held with representatives of the productive sector, academia, organizations, and national and international experts.
On that occasion, the meeting’s agenda focused on presenting initiatives aimed at sustainable economic development and technological innovation, with special emphasis on artificial intelligence, as well as alignment with the objectives of the Brazilian Artificial Intelligence Plan. At no point was there any request to direct speeches or any attempt to interfere with the lawmaker’s position.
It is important to stress that the visit did not influence Representative Vitor Lippi’s stance on the matter. The congressman has followed and studied the artificial intelligence agenda for many years, having engaged in dialogue with the OECD, specialists, academic institutions, and various national and international actors long before the emergence and popularization of tools such as OpenAI and others.
The lawmaker also notes that several companies, startups, and research and innovation centers in the sector have expressed concern about the direction of the legislation, particularly regarding the chapter on copyright.
The representative acknowledges and respects the concerns of the cultural sector and copyright holders, understanding that the issue can be properly addressed in specific legislation, such as the Copyright Law, or through a balance within the AI bill itself, so as not to harm any of the parties involved.
The statement made by the lawmaker before the Artificial Intelligence Commission reflects a consolidated, technical view that predates the aforementioned meeting, built through study, broad dialogue, and a public commitment to the country’s development. Vitor Lippi has always argued that Brazil must responsibly define what role it wants to play in the global AI landscape: merely as a user and spectator, or as a transformative agent, developer, and technology provider.
Excessively rigid regulation could make it unfeasible for innovative Brazilian companies to emerge and could also drive away strategic foreign investment. This is a risk the country cannot afford. Brazil needs to grow, create jobs, foster research and innovation, and not miss yet another historic window of opportunity in the field of disruptive technologies.
Statement by Jadyel Alencar
The office clarifies that dialogue with representatives of companies, entities, and organizations is a regular and republican practice of the mandate, just as it is with academia, the productive sector, civil society, and public authorities. Holding institutional meetings does not imply automatic alignment or influence over the lawmaker’s position.
With regard to Provisional Measure No. 1.318/2025, which establishes the Special Tax Regime for Data Center Services, it is important to note that the PM aims to encourage the installation and expansion of data centers in Brazil through tax incentives conditioned on counterparts such as investments in research and development, energy efficiency, and sustainability criteria. It is therefore a policy of economic, energy, and industrial development, not a measure aimed exclusively at so-called big tech companies.
It should also be clarified that Federal Representative Jadyel Alencar is not a member of the Joint Committee of Provisional Measure No. 1.318 of 2025 and therefore does not exercise any role as rapporteur or direct leadership in the work of the body responsible for analyzing the matter.
Representative Jadyel Alencar submitted three amendments with technical and structural content, aimed at legal certainty, fiscal control, and energy efficiency of the regime:
Amendment 132: enables direct connection between power generation units and energy-intensive consumers, including data centers, allowing off-grid or hybrid models. The amendment is part of the energy agenda and seeks to reduce costs, increase efficiency, and stimulate sustainable investments;
Amendment 133: establishes safeguards to ensure legal certainty and fiscal control in the temporary admission regime for imported goods, linking them directly to data center activities, with technical requirements defined by the Federal Revenue Service, the requirement of tax guarantees, and clear rules for conversion into permanent importation or re-export;
Amendment 134: reinforces the same axis of legal certainty and fiscal oversight, ensuring operational flexibility for companies without prejudice to inspection and tax collection.
Thus, the association of the amendments with specific interests of large digital platforms is surprising, since they address issues related to energy infrastructure, the tax regime, and fiscal governance, applicable to any enterprise classified under Redata.”
Statement by Luísa Canziani
Since I was elected, I have adopted respectful dialogue with everyone, without distinction, as a guiding principle of my mandates. Therefore, my office and my staff receive all those who seek us out: representatives from the productive sector, academia, public agencies, organized civil society, and workers. As chair of the Special Committee on Artificial Intelligence, we spoke with and listened to everyone who approached us, and I state that in none of my meetings, or those of my staff, was there any interference with my views or with the progress and direction of the committee’s work.
Regarding the mission to the United States, the travel costs were covered by the Parliamentary Front for the Free Market, which was responsible for organizing the official mission.
Ventura's office didn't answered until publication date.
This story was supported by a grant from the Tarbell Center for AI Journalism.
The reporting process
Aos Fatos compiled the names of lobbyists linked to the five main AI big tech companies and filed an Access to Information request with the House of Representative and the Senate to obtain records of their visits between the opening of the AI bill commission and the last public hearing on the topic. The data was tabulated and categorized by office visited, date, and company.
We then transcribed and analyzed lawmakers’ statements made both during hearings and at external events. We also investigated visits by lawmakers to big tech companies.
Finally, we gathered information on relationships between lawmakers and lobbyists or big tech companies.
