BRAZIL CONTEXT
Artificial intelligence for doctors in Brazil
Artificial intelligence for physicians in Brazil can support documentation, medical history organization, and evidence consultation. Its use must preserve the medical record and the physician's autonomy.
Brazil combines rules on record digitization, data protection, telehealth, and a specific CFM resolution for the use of AI in medicine.
Brazil Brazil
Central document: patient chart.
Digitization: Law No. 13,787/2018.
Health data: sensitive under the LGPD.
Medical AI: CFM Resolution 2,454/2026.
What a digital medical record must preserve in Brazil
The Law 13,787/2018 It regulates the digitization and use of computerized systems to keep, store, and manage medical records. The process must preserve integrity, authenticity, and confidentiality.
| Integrity | The digital document must reflect the clinical record without unauthorized alterations. |
| Authenticity | The institution must be able to demonstrate origin, authorship, and correspondence with the digitized or created document. |
| Confidentiality | Access must be restricted and protected in accordance with professional secrecy and data legislation. |
| Conservation | The law establishes rules for the archiving, digitization, and retention periods of the medical record. |
| Professional Review | The physician remains responsible for the clinical information that they validate and incorporate into the record. |
An automatic transcription or summary does not acquire the guarantees of the medical record on its own. It must undergo review and institutional system controls. The AI clinical documentation guide for Brazil offers an operational review.
FRAME FROM BRAZIL
Medical record, data protection, telehealth and AI
Brazil has distinct regulatory layers that complement each other. The institution must evaluate the clinical registry, data processing, care modality, and the purpose of the AI.
Law 13.787
It regulates the digitization, custody, and computerized management of patient medical records.
LGPD
Classify health data as sensitive and require adequate controls for its processing.
CFM 2,314
It regulates telemedicine and requires recording healthcare delivery in a physical medical record or electronic system.
CFM 2.454
Establish principles for AI in medicine, with human oversight, transparency, governance, and the physician's final decision.
University and high-complexity hospitals in Brazil
Brazil combines the SUS with large university, philanthropic, and private hospitals. Among the benchmark institutions for complexity, teaching, research, or innovation are:
Hospital of the Clinics of the Faculty of Medicine of USP
São Paulo. Public university complex of high complexity, teaching, and research.
Porto Alegre Clinical Hospital
Porto Alegre. Public university hospital linked to UFRGS and the EBSERH network.
Albert Einstein Israelite Hospital
São Paulo. Philanthropic hospital providing highly complex care, education, and health innovation.
Syrian-Lebanese Hospital
São Paulo. High-complexity philanthropic hospital with teaching and research activities.
Non-exhaustive selection of reference institutions. This is not a ranking nor does it indicate affiliation with Itaca.
Medical record, S-RES and RNDS: different functions
The local registry, the system that manages it, and the national exchange network are related, but they are not equivalent.
| Element | Main function | Responsibility |
|---|---|---|
| Medical record | Patient clinical information set and care record. | The establishment and the responsible professionals produce and safeguard it. |
| S-RES | Electronic Health Record System that captures and manages care information. | The organization defines security, profiles, signature, traceability, and integration. |
| RNDS | National Health Data Network for interoperability and continuity within the digital health ecosystem. | It does not replace the complete medical record kept by the provider. |
The CFM resolution keeps the final decision with the physician
CFM Resolution 2,454/2026 allows the use of AI as support for clinical decision-making, management, research, and education, but keeps the final say on diagnosis, treatment, and prognosis with the physician.
The standard requires human oversight and transparency toward the patient when AI plays a relevant role. It also addresses governance, risk classification, auditing, and data protection. The institution must document how it applies these principles to each tool.
Scope: A solution may also be subject to sanitary, medical device, telemedicine, professional ethics, and security requirements depending on its purpose.
Five questions to evaluate a tool
- Is AI classified according to its purpose and risk?
- Can the doctor accept, correct, or reject the discharge?
- Does the patient receive information when appropriate?
- Does the institution maintain governance, audit, and monitoring?
- Are the data processed in accordance with the LGPD and sanitary regulations?
Where AI can contribute to Brazilian clinical practice
At a continental scale system, AI contributes when it reduces repetitive drafting and facilitates continuity without replacing the medical record, the S-RES, or medical judgment.
Documentation
Drafting notes, reports, and summaries for the physician's review.
Standardization
Help apply shared structures and terminologies across services and locations.
Clinical evidence
Synthesize medical sources for specific questions and display references.
Continuity
Organize longitudinal background to retrieve relevant context.
AI does not solve interoperability, document quality, or data protection on its own. It is a support layer that must fit into the institution's controls and systems.
ITHACA IN BRAZIL
Itaca prepares clinical drafts. Your institution retains the official medical record.
Itaca helps turn conversations, documents, and quick notes into structured clinical drafts. It also allows working with templates and longitudinal context depending on the contracted functions.
The physician reviews and approves the content. The institution retains responsibility for the medical record, LGPD, access control, custody, and integration with its systems.
Official sources for further reading
- Law 13.787/2018: digitization and custody of medical records.
- Law 13.709/2018: General Data Protection Law.
- CFM Resolution 2,314/2022: telemedicine.
- CFM Resolution 2454/2026: Use of AI in medicine.
- Security and privacy in clinical notes with AI.
Editorial and source review: Itaca team. Updated: September 11, 2026. This content is informational and does not replace legal, regulatory, or security advice. Always verify the current version and the obligations applicable to your institution.
Frequently asked questions about medical AI in Brazil
Can AI make a diagnosis for the doctor in Brazil?
No. CFM Resolution 2,454 keeps the final diagnostic, therapeutic, and prognostic decision in the hands of the physician.
Does an AI tool replace the medical record?
No. It can prepare content, but the official medical record and its controls remain in the provider's system.
Are health data sensitive under the LGPD?
Yes. Its processing requires enhanced protection and an appropriate legal basis.
Must the teleconsultation be recorded?
Yes. CFM regulations require recording care in a physical medical record or in an Electronic Health Record System.
Explore other national contexts
Compare how the medical record, data protection, and interoperability change between countries.