Brazil: Introduced Bill No. 2338 of 2023 regulating the use of Artificial Intelligence, including algorithm design and technical standards

On 3 May 2023, the Bill No. 2338 of 2023 regulating the use of Artificial Intelligence, including algorithm design and technical standards, was introduced in the Senate of Brazil. In particular, the Bill would require providers of Artificial Intelligence systems to conduct a risk assessment to identify the level of risk it may pose before it is placed on the market. The providers would be required to include the purposes of the AI system and indicated applications in the preliminary assessment. The Bill differentiates between "excessive risk" and "high risk" AI systems. The "excessive risk" AI systems include systems that employ subliminal techniques that aim to or affect how individuals act that is harmful or dangerous to their health or safety, exploit the vulnerabilities of groups of individuals and systems that could be used by governments, to evaluate, classify or rank individuals and attribute a social score. The Bill specifies that a new authority designated by the executive branch will regulate the "excessive risk" AI systems. The "high risk" AI systems include systems used in critical infrastructure, professional education and training, recruitment, employment decisions, evaluation of criteria for accessing public goods, ranking priorities of emergency response services, administration of justice, autonomous vehicles, and biometric identification systems. The "high risk" AI systems would be required to meet the quality of service requirements and design their systems to allow for human oversight.