Previously Talisman AI
It was both an honor and a significant responsibility to address the Brazilian Senate on the matter of AI regulation. My speech was focused on the notion of “Brazilian dream” – meaning the ability of young, talented Brazilians to find a fruitful environment to start a business. I believe any AI regulation should have the purpose of advancing entrepreneurship.
I started off by sharing a brief overview of my journey in the United States, where the notion of American dream was born. At age 19, I pursued a master's degree at Harvard Law School, and it remains my proudest achievement that, as a Brazilian, I was recognized by the Wall Street Journal as the youngest person in decades to pass the New York Bar.
Despite having opportunities to settle abroad (including a work visa and permanent job offers in the United States), I returned home to embark on the "Brazilian dream" of starting a successful company, which led to the founding of Enter, an AI company committed to enhancing the efficiency of Brazilian enterprises such as Banco BMG, Sulamerica Seguros, Vivo, Nubank, Agibank, among others.
Secondly, I shared that the proposed AI regulation bill will harm AI innovation. Above all, while many legislators intend to limit the influence of Big Tech giants, I emphasized that "regulating artificial intelligence will primarily harm AI startups," not just the Googles and Metas of the world.
This legislation proposal, particularly Articles 17 and 50, impose preemptive government censorship on AI models even before they scale to broad use and threaten significant fines (the largest of $10 million dollars or 2% of the company's annual revenue) as punishment for violating a bureaucratic procedure that will be created to submit AI models to review by a government commission. This heavy-handed approach would hinder our access to advanced models from global leaders like OpenAI and Google, as a result of the chilling effect of releasing models in Brazil, effectively making Enter less competitive on the global stage.
In conclusion, while I recognize the necessity for responsible AI deployment, I urged the Senate to consider "maturing the debate around AI before implementing regulation." Existing legislation on tort law and consumer protection is sufficient to ensure that bad actors are punished while allowing innovative companies to build AI models freely. That is how the world's largest AI builders – China and the United States – have chosen to approach the matter.
We must not stall the growth of startups that will drive Brazil's advancement in AI technology. Brazil must champion innovation, not burden it by premature regulatory structures. As I emphasized to the senators, "because the constitutional mission of the Federal Senate is to ponder legislation with consideration and restraint, the best path for this bill may be for it to wait patiently."
