In a debate at the Web Summit Rio, Ronaldo Lemos and Bruno Lewicki defended legislation that protects rights, but also encourages innovation, development and the creation of national capabilities in artificial intelligence. Raoni Alves / g1 Rio While the Chamber of Deputies discusses the regulation of artificial intelligence in Brazil, experts criticized the model that has been built and defended that the country develop its own strategy for the sector, aimed not only at mitigating risks, but also at technological development and reducing inequalities. Earlier, this Tuesday (9), the president of the House, deputy Hugo Motta (Republicanos-PB) ruled out that the vote would take place this week. 📱Download the g1 app to see news from RJ in real time and for free The statements were made this Tuesday (9), during the panel "Can we open AI?", at the Web Summit Rio, one of the main technology events in Latin America. The debate brought together Ronaldo Lemos, founding director of the Institute of Technology and Society (ITS-UERJ) and one of the creators of the Marco Civil da Internet, and Bruno Lewicki, head of Public Policies for Latin America at OpenAI, the company that created ChatGPT. Although they have different perspectives on some aspects of regulation, both converged in the assessment that Brazil needs to balance protecting rights and encouraging innovation, avoiding creating an environment that hinders the development of artificial intelligence in the country. Web Summit debates AI and technology at Riocentro Brazilian model of regulation The most scathing criticism came from Ronaldo Lemos of the model that, according to him, has been discussed in the National Congress. For the expert, the proposal approved by the Senate in 2024 would have excessively followed European legislation, precisely at a time when the model adopted by the European Union itself was already being reviewed. "This is very uncomfortable for me, especially because the Brazilian law is being discussed in Congress, in its latest version that came from the Senate, it copies the European law. This bothers me deeply for two reasons. First, because it copied the European law of 2019. And since 2019, the European law has already been completely modified", explained Ronaldo. In the expert's assessment, Brazil has enough experience to develop its own legislation. "We don't need to copy anyone. We can create a Brazilian law on artificial intelligence, and this law can be an example for the whole world", he argued. According to him, the focus should be on a policy that combines protection of rights, training of the population and strengthening of national technological capabilities. Lack of transparency Another point of criticism was the process of preparing the proposal. When comparing the current debate with the construction of the Marco Civil da Internet, Ronaldo stated that he felt the lack of a public and participatory process. "The Marco Civil is a law made through a participatory process (...). Perhaps it was the only Brazilian law written in plain sight, openly, on the internet. I miss that a lot", he commented. Web Summit Rio 2026 begins with a focus on AI, creators and the global dispute over technology Reproduction Next, he made the most direct criticism of the progress of AI regulation. "Everyone is saying: 'Oh, you're going to vote next week', but no one has seen the text. We are about to have a vote on a text that no one has seen." The statements come on the same day that Hugo Motta stated that the report by deputy Aguinaldo Ribeiro (PP-PB) is still being prepared and that its processing depends on alignment with the Senate. Protection of rights and stimulation of development Representing OpenAI at the event, Bruno Lewicki defended the need for balanced regulation. According to him, the objective must be to protect fundamental rights without impeding technological advancement and innovation. "Speaking specifically about regulation of technology, internet, I think the main objective should be to strike a balance. You need to protect citizens' fundamental rights from risks that are exacerbated, but you also need to stimulate, not only investment, but stimulate development", commented Bruno. For him, the debate should be part of a broader strategy for the country. "I think that, fundamentally, what we should be discussing is a country project and how the country intends to incorporate artificial intelligence into this project." Experts criticize AI regulation model in Brazil Raoni Alves / g1 Rio Lewicki also stated that, based on the text approved by the Senate, he perceives a greater concern with restrictions than with incentives. "It looks a bit like a football team that was fielded with eight or nine defenders and one attacker", he analyzed. In the executive's assessment, unbalanced legislation can have consequences for national development itself. "Legislation that is unbalanced generates a paralyzing effect not only for foreign investments, but for the development of the country itself", he commented. Risk of technological dependence Ronaldo Lemos also stated that Brazil's main challenge is not just regulating artificial intelligence, but developing its own capacity to produce it. "In my view, the biggest risk that Brazil runs is that we do not develop any capacity to have artificial intelligence in Brazil, that we create a total dependence on external artificial intelligence." He cited countries such as Vietnam and India as examples of nations that also use regulatory policy as an instrument of economic and industrial development. For the expert, the discussion about AI needs to incorporate topics such as education and inequality. "We have to think about AI regulation in all its complexities: protection against inequality, training." Web Summit Rio 2026 begins with a focus on AI, creators and the global dispute over technology Reproduction He then drew attention to something that he considers worrying given the expansion of technology. "The main requirement to use artificial intelligence well is mastery of the Portuguese language." According to him, as only a small portion of the population fully masters the language, there is a risk that artificial intelligence will increase existing inequalities. "If we let (...) let go and leave things as they are, what I think could happen in 10 years is an increase in inequality", assessed Lemos. What the text approved by the Senate provides The artificial intelligence regulation project was approved by the Senate in 2024 and still needs to pass the Chamber of Deputies before proceeding for sanction. The proposal establishes rules for the commercial use of AI in the country, provides transparency in the identification of artificial intelligence systems, creates specific obligations for generative AI tools and establishes sanctions for companies that fail to comply with the legislation, including fines that can reach R$50 million or 2% of revenue per infraction, in addition to other administrative penalties. The project's objectives include defining limits and permissions for technology applications in areas such as public services, infrastructure and security, while also seeking to establish accountability mechanisms for the use of artificial intelligence in Brazil. READ ALSO: Government establishes new rules for big techs to operate in Brazil; see main points Internet civil milestone: how the US and Europe treat 'big techs' Brazil, France and India unite in defense of global regulation of artificial intelligence