Brazil Senate bill PL 931/2026 would set guidelines for AI in healthcare and was awaiting a committee vote on 14 October 2026

Key takeaways

• The bill is pending. Agência Senado says it "aguarda votação na Comissão de Assuntos Sociais (CAS)".
• The report says the bill provides that AI use in care "não afasta a responsabilidade dos profissionais de saúde nem substitui a supervisão humana".
• Health service users would have a right to clear and adequate information about AI use in their care, the report says.
• Per the report, the substitute avoids overlapping topics beyond the health-AI guidelines addressed in the bill, since much general AI governance is already covered by PL 2.338/2023, which is with the Chamber of Deputies.

Is PL 931/2026 law, and where is it in the Senate?

Based on the dated Senate snapshot, it was not law as of 14 October 2026. Agência Senado, the Senate's news service, reported on 14 October 2026 that the bill by Senator Dra. Eudócia (PSDB-AL) is reported by Senator Damares Alves (Republicanos-DF), who changed the original text, and that it "aguarda votação na Comissão de Assuntos Sociais (CAS)". The report gives no vote date. The 14 October report is a snapshot, and we did not check the Senate bill-tracking page, so the current stage may differ. Passing the CAS would not by itself make it law, and the report does not describe the later steps. Our description of the bill comes from this news report, not from the bill text.

What would the bill require for AI in healthcare?

Agência Senado says AI use in health would observe patient protection and safety, transparency and the preservation of health professionals' autonomy, among other foundations. The report states that the bill provides that using AI in care activities "não afasta a responsabilidade dos profissionais de saúde nem substitui a supervisão humana". It also says users of health services would have the "direito à informação clara e adequada sobre a utilização de sistemas de inteligência artificial em seu atendimento". Where AI is used inside the public health system (SUS), the report says use would follow, where applicable, the standards set for the national health data network.

On oversight, the report says regulation and inspection of AI applications would be carried out through coordinated action by the bodies responsible for health public policy, health regulation and personal data protection. The report names no agency and mentions no penalties or deadlines, and it specifies no AI-specific technical requirements beyond the SUS data-network standards noted above.

How does it relate to Brazil's general AI bill, PL 2.338/2023?

The rapporteur's report observes that much of general AI governance is already addressed in PL 2.338/2023, by Senator Rodrigo Pacheco, which the Senate approved in December 2024 and sent to the Chamber of Deputies. Agência Senado says the rapporteur still saw merit in the health proposal and, to avoid overlap, her substitute text keeps to health-sector safeguards, aimed at protecting patients and preserving human supervision in higher-impact clinical activity. Our separate briefing on the status of PL 2338 covers the general bill. Professional-conduct rules already exist for physicians, as covered in our note on the CFM resolution on physician responsibility for AI, a separate instrument.

What does it mean for healthcare counsel and health-tech teams?

The practical signal is about allocation of responsibility. The report says the bill provides that AI use in care does not remove the responsibility of health professionals, which is the point malpractice counsel and insurers would watch. The reported information right is a right of health service users; the report does not say who must provide it or whether vendors have any duty. Whether vendors would be asked to support it through product design or documentation is our speculation, not something the report states. These are our readings of the reported text, and the final wording could differ after the CAS vote.

For a comparable health-sector rule elsewhere in the region, see our note on Colombia Resolution 1644 on telemedicine and AI supervision, which is a different jurisdiction and instrument.

What should teams do now?

• Add PL 931/2026 to your Brazil watch list and obtain and read the substitute text, which the report describes as already presented, and track the CAS decision. We did not check whether the substitute text is publicly available.
• Map which clinical AI tools you use or sell in Brazil and whether a human reviews their output before it affects care.
• Check whether patient-facing materials would let you tell a patient that AI was used, since the report says health service users would have a right to that information.
• Do not describe the bill to clients as an obligation. Describe it as a pending proposal.

What are the key dates and parties?

What we did not verify

What we opened: the Agência Senado news report of 14 October 2026 (09h04) on the Senate website, on 2026-10-25.

What we did not open: the text of PL 931/2026 as filed, the rapporteur's substitute text and report, the Senate bill-tracking page, PL 2.338/2023, and any committee agenda. We cannot say when the CAS will vote, whether it will adopt the substitute, or what later changes occur. The quotations are from the news report's own wording about the bill, not from the bill text.

What we refuse to claim: that the bill is law, that it has a particular effective date, that it creates penalties or sets a regulator, or what the final text will say.

Related briefings

• AI Regulation News hub
• Latin America AI regulation tracker
• AI in healthcare compliance
• Brazil PL 2338 AI framework: status
• Brazil CFM resolution on physician responsibility for AI
• Colombia Resolution 1644 on telemedicine and AI supervision