On April 24, 2024, the Senate announced that the Temporary Commission on Artificial Intelligence in Brazil (CTIA) presented its preliminary report on the regulation of artificial intelligence (AI), containing proposals for an AI bill. The preliminary report aimed at being aligned with existing regulations in the EU and the US.

What is the scope of the preliminary report?

The proposals contained in the preliminary report would apply to the conception, development, implementation, use, adoption, and responsible governance of AI systems. Exceptions include AI systems that are:

• used by individuals for non-commercial activities;
• developed and used exclusively for national defense purposes; and
• used for testing, research, and development, where not made available on the market.

What are the main aspects of the preliminary report?

The preliminary report contains proposals aimed at developing and using AI systems that, among other things:

• are centralized around the human person;
• respect human rights and democracy;
• protect the environment;
• do not discriminate; and
• promote economic, scientific, and technological development and innovation.

Rights

The CTIA also proposed rights for people and groups affected by AI systems, namely the:

• right to information;
• right to privacy and data protection;
• right to determination and human input in decisions made by AI systems; and
• right to not be discriminated against and the right to correct discriminatory acts.

The preliminary report also provides specific rights to people and groups affected by AI that produce relevant legal consequences or high-risk AI.

Risks

According to the preliminary report, agents implementing AI systems would have to conduct a risk assessment before making the AI system available on the market. The preliminary report also contains a list of criteria to determine if an AI system represents excessive risk, and is therefore prohibited. The implementation of high-risk AI systems would need specific regulation.

Other provisions

Additionally, the preliminary report contains provisions regarding:

• AI systems governance;
• civil liability of providers and operators of AI systems;
• certification schemes;
• notification of security incidents;
• supervision by the National System of Regulation and Governance of Artificial Intelligence (SIA), created therein;
• copyright; and
• sanctions for non-compliance.

You can access the preliminary report here and the press release here, both only available in Portuguese.