Brazil - National AI Committee (270/2025)

Portaria CNJ No. 270/2025

Portaria CNJ nº 270/2025

Brazil

RAI-BR-NA-PCN2PXX-2025
Effective: August 27, 2025
In Force(In Force)
RegulationGovernance and OversightRisk Management
Export PDF

Portaria CNJ nº 270/2025 designates the members of the National Committee on Artificial Intelligence of the Judiciary (CNIAJ), the CNJ body tasked with assisting the Conselho Nacional de Justiça in implementing, supervising and monitoring the application of the CNJ’s AI governance Resolution (Resolução CNJ nº 615/2025). It establishes composition, secretariat, basic functions and reporting obligations for the Committee and integrates with Resolução 615/2025 governance, transparency and audit requirements.

Overview

Portaria CNJ nº 270/2025 designates the members of the Comitê Nacional de Inteligência Artificial do Judiciário (CNIAJ), the multidisciplinary oversight body created by Resolução CNJ nº 615/2025. The Portaria, dated 27 August 2025 and published in the CNJ Electronic Journal (DJE/CNJ) on 3 September 2025, operationalizes the Resolution by naming titular and alternate representatives from the CNJ, magistrature, Public Ministry, defenders, the Bar Association (OAB), academia, civil society and technical staff for an Executive Secretariat. The Committee’s mandate is strategic and advisory: it supports implementation, supervision and harmonization of AI governance across the Brazilian Judiciary, with responsibilities ranging from risk reclassification and audit definition to technical guidance for the Sinapses platform and public reporting obligations. The text of the Portaria is published on the CNJ Atos Normativos system (Portaria No. 270/2025) and summarized on the CNJ portal (CNJ CNIAJ page).

Definitions

For purposes of the Portaria and its implementing framework, key terms follow the definitions in Resolução 615/2025: "AI solution" (any system developed or deployed using machine learning, statistical inference or other automated decision-support technologies), "risk classification" (categorization of systems as low, high or excessive risk with differing governance obligations), "audit" (technical and procedural review of AI systems to verify compliance with governance, privacy and fairness requirements), "transparency report/impact report" (public documents describing the functioning, training data categories, risk mitigation and human oversight mechanisms of AI systems), and "Sinapses" (the CNJ’s centralized platform/registry for AI systems and projects in the Judiciary). The Portaria uses these established terms to anchor the Committee’s remit while delegating technical detail to the Resolution and subsequent Committee-delivered norms.

Governance and Institutional Framework

The Portaria sets out the CNIAJ composition (titular and alternate members) drawn from multiple institutional actors (CNJ counselors, technical magistrates, representatives of the Public Ministry, public defenders, the OAB, technical secretariats, academia and civil society). The Act designates an Executive Secretariat (a judge and a CNJ staff member) to provide administrative support and link the Committee with the CNJ’s Presidency. The composition ensures plural representation and aligns with the governance model required by Resolução 615/2025, which contemplates interdisciplinary oversight (art. 15–16 of the Resolution). The Committee is institutionally positioned to produce annual work plans, meeting minutes and an annual report of activities for publication on the CNJ portal; it may propose technical standards and cooperation agreements and recommend audit activities and capacity-building measures for tribunals. The Portaria thus operationalizes an oversight architecture that centralizes strategic direction at CNJ while preserving operational responsibilities at the tribunal level under the supervision framework of the Resolution.

Key Focus Areas

The Committee’s principal focus areas—expressly connected to the duties assigned by the Resolution and operationalized by the Portaria—include: (1) risk classification management (review and reclassification of AI systems used or proposed by tribunals); (2) transparency and reporting (establishing templates and schedules for public impact reports and non-technical disclosures to be posted on Sinapses); (3) audit and evaluation frameworks (defining technical standards, periodicity, and the scope of algorithmic audits for high-risk systems); (4) data protection and cybersecurity alignment (ensuring LGPD compliance and secure handling of datasets used to train models); (5) capacity building (coordinating with Enfam and Enamat to establish curricula and training for magistrates and servers); (6) procurement and market assessment (assessing external suppliers, recommending adoption or suspension of market solutions based on security, privacy and fairness criteria); and (7) international cooperation and standards alignment (consolidating best practices and formal cooperation with foreign and multilateral bodies). These focus areas reflect the CNJ’s intent to harmonize the Judiciary’s AI activities inside a rights-respecting and accountable governance framework, and they are consistent with the Committee objectives listed on the CNJ information pages (CNIAJ page).

Implementation Framework

Implementation is a two-tiered approach: the Portaria names the governing body and Secretariat that coordinate nationally, while operational detail and compliance remain the obligation of individual tribunals, which must register AI projects on the Sinapses platform and follow the Classification and audit processes established by Resolução 615/2025. The Committee issues guidance documents, templates for impact and transparency reporting, and technical audit protocols; it may request reclassification or order audits for systems initially classified by tribunals. Coordination with existing regulatory instruments—such as the LGPD (Lei nº 13.709/2018) and CNJ contracting/procurement rules—will be essential to ensure contract clauses, data processing agreements, and security measures conform with judicial standards. The Portaria thus focuses on governance design and delegated authority while relying on the Resolution and Committee outputs for operational rules.

Monitoring and Evaluation

The Portaria requires the Committee to produce annual workplans and reports and to maintain minutes of meetings; these transparency obligations enable external monitoring by the CNJ and by interested stakeholders. The Committee’s monitoring toolkit includes periodic technical audits (mandated for high-risk systems by the Resolution), maintenance of a public catalog of AI systems in Sinapses, and development of risk reassessment triggers (e.g., material changes in model training data, performance degradation or reports of discriminatory outcomes). The Committee is also empowered to recommend targeted audits or suspensions of systems that fail to meet governance standards. Monitoring will be complemented by metrics and KPIs produced by the Committee and CNJ technical teams to track compliance, number of registered systems, number of audits concluded and remediation measures implemented.

Penalties, Liability, and Appeals

The Portaria itself is an administrative designation and does not prescribe independent penalties. Enforcement and sanction regimes derive from Resolução 615/2025 and existing CNJ disciplinary and administrative instruments: non-compliance can trigger remedial orders, mandatory audits, suspension of system use, corrective plans, reporting to responsible procurement units, and, where appropriate, referral for disciplinary proceedings under applicable judicial administrative rules. Liability for harms caused by improperly governed AI systems remains governed by Brazilian civil and administrative law, and impacted individuals retain access to remedies under the LGPD and ordinary judicial procedures. The Committee’s role includes recommending corrective and enforcement measures to the CNJ based on audit outcomes and risk assessments.

Relationship to Other Instruments

The Portaria implements and operationalizes the institutional elements of Resolução CNJ nº 615/2025 (the substantive AI governance framework). It interacts with prior CNJ norms (e.g., Resolução CNJ nº 332/2020) and with national legal regimes such as the LGPD (Lei nº 13.709/2018), rules on access to information, procurement law, and judicial disciplinary frameworks (Lei Complementar nº 35/1979 and others). The Committee is also expected to coordinate with judicial training schools (Enfam/Enamat) and with technical security bodies inside the Judiciary to ensure alignment between governance, staffing and technological safeguards. When transnational or sectoral questions arise (e.g., use of foreign-sourced AI models), the Committee’s guidance will be informed by international best practices and by sector-specific standards applicable to health or financial data when used in judicial contexts.

International Alignment

The CNIAJ’s formation reflects global trends toward sectoral AI governance and aligns conceptually with international instruments (including elements similar to the EU AI Act’s risk-based approach and recommendations from multilateral organizations). The Portaria positions the Committee to pursue international cooperation and to integrate international standards into CNJ guidance. The Committee may therefore engage in bilateral or multilateral cooperation, participate in international fora, and adapt foreign technical standards for the Brazilian judicial context while ensuring compatibility with domestic law, notably the LGPD. The CNJ portal and press releases indicate an intention to consolidate dialogue and cooperation with external institutions (CNIAJ page).

Implementation Timeline

EventDate
Portaria signed2025-08-27
Portaria first published in DJE/CNJ2025-09-03
Republication to correct material error2025-09-12
Committee begins operational work publicly (first meetings reported)2025-09-30

Compliance Checklist

RequirementWhoNotes
Register AI projects on SinapsesTribunals/UnitsPer Resolução 615/2025
Submit high-risk systems to auditTribunals/Service ProvidersAudit protocols to be defined by CNIAJ
Publish transparency/impact reportsResponsible TribunalNon-technical and technical versions
Comply with LGPD and security standardsAll unitsData processing and contracts
Cooperate with CNIAJ requestsTribunalsProvide documentation and access for audits

Sources and References

SourceType
Portaria Nº 270/2025 (designation of CNIAJ members) - Atos CNJPrimary Source
CNJ - Comitê Nacional de Inteligência Artificial do Judiciário (CNIAJ) - Portal CNJPrimary Source
Resolução CNJ Nº 615/2025 (AI governance in the Judiciary) - Atos CNJPrimary Source
Plain English

Brazil's National Council of Justice (CNJ) has established a new committee to oversee the use of Artificial Intelligence (AI) across the country's judiciary, impacting all courts and legal professionals developing or deploying AI solutions.

This administrative order, Portaria CNJ nº 270/2025, officially designates the members of the National Committee on Artificial Intelligence of the Judiciary (CNIAJ). This multidisciplinary body brings together representatives from the CNJ itself, judges, public prosecutors, public defenders, the Brazilian Bar Association, academia, and civil society. Its primary role is to assist the CNJ in implementing, supervising, and monitoring the application of the broader AI governance framework outlined in Resolution CNJ nº 615/2025.

For product managers and in-house teams, the CNIAJ will be the central authority defining how AI is developed and used within the Brazilian Judiciary. Its key functions include: - Reviewing and reclassifying AI systems based on their risk level (low, high, or excessive). - Establishing templates and schedules for public transparency and impact reports, which must detail how AI systems function, their training data, and risk mitigation. - Defining technical standards and periodicity for algorithmic audits, especially for high-risk systems. - Ensuring all AI solutions comply with Brazil's General Data Protection Law (LGPD) and cybersecurity standards.

The practical implication is that all courts and judicial units must register their AI projects on the CNJ’s Sinapses platform and adhere to the guidance and audit requirements issued by the CNIAJ. This Portaria was signed on August 27, 2025, and the Committee began its operational work by September 30, 2025.

While this Portaria itself doesn't prescribe penalties, non-compliance with the CNIAJ's directives or the underlying Resolution 615/2025 can lead to significant consequences. These may include mandatory audits, suspension of AI system use, corrective action plans, and potential disciplinary proceedings under existing judicial administrative rules. A crucial pitfall to remember is that the CNIAJ is the *watchdog* that will translate the general AI governance Resolution into concrete, enforceable rules. Therefore, understanding the Committee's guidance will be paramount for any team deploying AI in the Brazilian Judiciary.

Plain-English rewrite by Regulations.ai — not legal advice. Verify against the official text.

What you must do — compliance checklist

0 / 5 marked complete

Plain-English obligations under Brazil - National AI Committee (270/2025). Not legal advice — verify against the official text before relying on it.

  1. #1CriticalImplementation Framework

    Applies to: Individual tribunals and judicial units

    operational detail and compliance remain the obligation of individual tribunals, which must register AI projects on the Sinapses platform
  2. #2CriticalMonitoring and Evaluation

    Applies to: Tribunals and service providers of high-risk AI systems

    periodic technical audits (mandated for high-risk systems by the Resolution)
  3. #3CriticalKey Focus Areas

    Applies to: Responsible tribunals deploying AI systems

    establishing templates and schedules for public impact reports and non-technical disclosures to be posted on Sinapses
  4. #4CriticalKey Focus Areas

    Applies to: All judicial units and AI system providers

    ensuring LGPD compliance and secure handling of datasets used to train models
  5. #5CriticalImplementation Framework

    Applies to: Individual tribunals and judicial units

    it may request reclassification or order audits for systems initially classified by tribunals

© Regulations.AI — created on 13-Jun-2026