What does Brazil want from artificial intelligence? – 04/07/2024 – Ronaldo Lemos

The Senate announced that it wants to approve a bill regulating artificial intelligence by the end of April. This statement is astonishing. There is still no consensus on the topic. There is agreement regarding their problems, but no one knows for sure which remedies to apply. In this uncertain scenario, it is clear that society should be closer to the challenge of regulating artificial intelligence.

To begin with, the project being analyzed in the Senate has a fundamental problem. It is inspired by the legislative model developed in Europe, the so-called AI Act (Artificial Intelligence Law there). Is what is good for Europe good for Brazil? This assumes that the medicine created there would work here. There are those who say that, as it was unable to export technology, the region then began to export legislation on technology.

Another problem is that society has so far been left out of the regulatory debate. The various economic sectors, academia, civil society or the scientific community and people in general have not yet been able to contribute to the text. The consultations carried out were not sufficient. When writing the final texts, the lawyers’ pen predominates.

It is worth remembering that the European AI law has not even been tested, as it will only come into force in its entirety in 2026. If the Senate votes on a law similar to it, we will have a curious case of premonitory importation: the adoption of a law whose effects are unknown even in their place of origin.

To regulate artificial intelligence, it is necessary to first answer: “What does Brazil want from artificial intelligence?” All regulation must be based on the answer to this question. And the answer must come from society. Only through broad participation will we be able to deal with the complexity of artificial intelligence, as it affects all of our lives.

Brazil certainly wants protection from the problems that AI can bring. Therefore, the Superior Electoral Court has already regulated the most urgent point: the impact of AI on the 2024 elections, including deepfakes. By resolving this immediate action point, the TSE gave us the necessary time to address other structuring issues, including the country’s strategy and competitiveness in the face of AI. These are matters to be considered calmly. There is no rush that is justified at this time.

Furthermore, the Federal Court of Auditors published an important report on Friday in which it harshly criticizes the bill currently underway in the Senate. In the words of the TCU: “The project emphasizes the reckless aspect of AI in the country, to the detriment of the adoption of guidelines capable of promoting the responsible development of technology. Disproportionate regulation may result in excessive expenditure of time and money to comply with complex rules , instead of promoting investment in technological development”.

It will be a disappointment if the path to regulating AI among us is derived from a model made elsewhere. It will also be disappointing if society does not truly participate in this process. It is worth remembering that the European law itself took four years to be drafted, assisted by several society committees. The USA has broad public consultation to regulate the issue. Taiwan has created an ambitious process to listen to the entire population to regulate AI. Brazil has also done this with the Marco Civil. It is time for society to resume its role in regulating this, which is one of the most important issues of the present.

READER

It’s over – Regulate artificial intelligence without broader contributions from society

Already – Countries that dare to forge their own path in AI regulation

It’s coming – Brazilian society organizing to participate in AI regulation

LINK PRESENT: Did you like this text? Subscribers can access five free accesses from any link per day. Just click the blue F below.