Skip to main content
v2026.11,986 entries · CC-BY 4.0

Editorial · CASRAI · AI and ML research outputs

Brazil adopts AI education rules covering university governance

Brazil’s AI education resolution requires research disclosure and university policies, with a 12-month transition and immediate prohibitions.

Published 16 Nov 2026· 5 minute read

Ask CASRAI · free to try

Ask about this story

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

CASRAI is the reference for research administration — bookmark it for the next question.

Brazil’s national AI education rules took effect on November 12, 2026, bringing public and private universities within a framework for research disclosure, institutional governance and human oversight. The National Education Council’s Resolution CNE/CP 1/2026 gives institutions 12 months to adapt policies and contracts, while applying its Article 41 prohibitions immediately.

From ministerial ratification to binding resolution

The legal sequence matters. Education Minister Leonardo Barchini ratified Opinion CNE/CP 15/2026 through a dispatch dated November 10 and published in the official gazette on November 11. The resulting resolution was signed on November 11. Its gazette publication followed on November 12, in edition 213-B, extra section 1. Article 47 makes publication the effective date.

The resolution covers educational institutions and systems across all levels and modalities. For higher education, Article 25 permits AI both as an object of study and as support for teaching, research and university extension activities, subject to course characteristics and institutional educational plans. The framework therefore reaches beyond classroom chatbot use into the organization of academic work.

Relevant AI contributions to research must be declared

Article 28 of the official resolution requires declaration of AI use whenever its contribution is relevant to writing, data analysis, literature review, programming or other stages of scientific investigation. It ties that obligation to scientific integrity, methodological transparency and intellectual responsibility. Authorship and responsibility for results remain entirely with the researchers.

The relevance condition is significant: the provision addresses the contribution to the research process, rather than listing a single prohibited tool or treating every interaction identically. Article 10 separately assigns students, faculty and researchers responsibility for selecting and verifying information and for its truthfulness, consistency and suitability. AI assistance does not transfer those responsibilities to the technology.

For research offices, the practical implication is that disclosure arrangements sit alongside human verification and authorship policy. The resolution sets a substantive obligation; Article 27 allows institutional policies to address appropriate forms of declaration. A generative-AI disclosure statement explains the underlying transparency concept, while the Brazilian rule expressly extends to relevant contributions across multiple research stages.

University governance extends to policies and procurement

Article 27 requires higher education institutions to establish their own AI policies for teaching, research, extension and academic management. Where pertinent, those policies must be incorporated into institutional development plans, known as PDI, course pedagogical projects, known as PPC, institutional bylaws, academic regulations and other governance instruments. This places AI within existing institutional decision structures.

The general governance provisions add operational requirements. Article 12 calls for institutional records of adoption and use decisions, periodic policy review and faculty participation in oversight. Where education professionals conduct human review, institutions must provide appropriate training, institutional time, technical support, consultation channels and decision-making backing.

Procurement is also covered. Article 5 requires contracting instruments to secure governance, auditing, monitoring and accountability mechanisms directed at suppliers and other responsible technology providers. Article 6 requires a reasoned technical, pedagogical and institutional adoption decision, with community participation and consideration of evidence, infrastructure, data protection and institutional capacity. These provisions connect AI governance to the terms under which a university acquires and supervises a system.

Human assessment cannot become a rubber stamp

The final text draws a precise boundary around assessment. Article 40 classifies automated pre-correction of essays and other work requiring interpretive judgment as high risk, but permits it as preparatory assistance to a teacher’s correction. Under paragraph 4, AI-generated scores or judgments cannot have academic effect without human correction. The assessor must read and directly appraise the entire student submission, with a right to human review.

Tacit approval, validation in batches and mere ratification of the system’s suggested result are prohibited. For objective assessments, AI may also serve as an auxiliary instrument, but academic consequences require effective, prior, qualified and documented human validation. These distinctions mean the resolution permits specific forms of assistance under safeguards.

Article 41 separately prohibits academic or disciplinary sanctions based solely on an automated AI-content detector. That boundary is directly relevant to university academic integrity policies: detector output cannot by itself determine a sanction. Article 29 also prohibits exclusively automated decisions with a relevant impact on students’ rights or academic trajectories.

Risk controls and the response from higher education

Articles 36–41 establish four risk levels, assessed according to purpose, autonomy, data and effects on students. High-risk applications require reinforced safeguards, including prior algorithmic impact assessment, documented human supervision, auditability and effective challenge and review. A personal-data protection impact report is required where applicable. Institutions must periodically reassess classifications and safeguards as uses change.

Independent reporting captured the policy’s human-responsibility emphasis before publication of the resolution. SBT News reported on November 10 that relevant research contributions would need declaration and that researchers would remain responsible. In Folha’s November 10 reporting, co-rapporteur Celso Niskier described negotiations over assessment wording and emphasized the teacher’s final responsibility.

After the resolution’s publication, the private higher education association ANUP said on November 13 that it would analyze the text’s effects on academic activities and member institutions’ management. Its notice highlighted institutional policies, research disclosure and the distinction between the adaptation period and immediate prohibitions.

A 12-month transition with immediate prohibitions

Article 46 starts the 12-month adaptation period from publication on November 12. It covers policies, practices, governance instruments and existing contracts, expressly preserving immediate application of Article 41. Universities therefore face both institutional implementation work over the transition and prohibitions already in force.

Article 45 assigns the CNE ongoing implementation monitoring in coordination with education systems, institutions, the scientific community and other participants. It provides for revisions when relevant technological, scientific, pedagogical or regulatory changes justify them. For university research offices, the concrete development is an enacted education resolution that connects research disclosure to institutional policy and oversight, with a defined transition rather than a future proposal awaiting approval.

Sources

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →