The bill regulating artificial intelligence (AI) in Brazil was approved by the Senate on Tuesday. It establishes rules for the development and use of AI in the country, ensuring protection of rights in areas such as work, security, health, and others.
The AI systems will be classified into risk levels based on their impact on human life and fundamental rights, with specific rules for each type. Low-risk tools will adhere to general data protection and fundamental rights norms, such as virtual assistants, translators, and public service systems.
High-risk initiatives will undergo algorithmic impact assessments, including autonomous vehicles, water and electricity supply projects, selection processes like exams, investigation and security services, health, and education.
The new legislation prohibits excessively risky systems, such as autonomous weapons and mechanisms using subliminal techniques to influence human behavior or exploit vulnerabilities to induce harmful behaviors or unequal access.