BRAZIL Trends and Developments Contributed by: Eduardo Perazza, Glaucia Coelho, Débora Chaves Martines Fernandes and Agnes Domingues, Machado Meyer
tion of data protection rights and freedom of expres- sion in the context of AI-generated content, where no settled precedent exists and the legal framework itself is in flux. Or a complex commercial dispute turning on the interpretation of a hybrid contractual arrangement that spans multiple regulatory regimes, where the admissibility of the special appeal depends on dem- onstrating a conflict of interpretation among appellate courts, a conflict that may be emerging but has not yet crystallised into a cleanly identifiable split. These are not hypothetical scenarios; they are the types of cases that regularly arise in sophisticated litigation. The very logic of AI, based on patterns extracted from historical data, finds its limits precisely where strategic litigation begins: in matters that require the construction of new legal arguments rather than the application of existing ones, and where the strength of the case lies not in its conformity to established patterns but in its capacity to persuade the court to chart a new course. AI and arbitration: a parallel transformation If the cases that most resist algorithmic treatment are those demanding original argumentation, bespoke strategy and deep subject-matter expertise, then international arbitration is, by vocation, the terrain where the limits of AI should be most acutely felt. Arbitration exists precisely because certain disputes require a procedural framework that standard litigation cannot offer. In Brazil, this is not a niche market: 376 new arbitrations were filed in 2024, an 18% increase over the previous year, involving BRL76 billion in dis- pute. Yet for all its scale and sophistication, the Bra- zilian arbitral market is only beginning to reckon with AI in the conduct of proceedings. Institutions have invested in digital infrastructure – CAM-B3 was among the first to adopt a fully digital case management platform – but the adoption of AI in the strict sense remains nascent. The most significant initiative to date came from the CAM-CCBC, which in October 2025 launched, in partnership with Jus Mundi, AI-generated summaries of arbitral awards, reviewed by specialised professionals before publication, the first application of AI for transparency in Brazilian arbitration. These initiatives, however, remain focused on institutional transparency rather than on the use of AI within pro- ceedings themselves.
As the Queen Mary University of London 2025 Inter- national Arbitration Survey documented, AI adoption across the arbitral ecosystem has accelerated to the point where institutional responses have become una- voidable. International bodies have moved accord- ingly: the CIArb published detailed guidelines, the SCC issued formal guidance in October 2024, the VIAC followed in April 2025, and the ICC established a dedicated Task Force in 2025. Among Brazilian insti- tutions, however, only the CAM-CCBC has taken a formal step issuing guidance that alerts participants to risks of imprecision, bias and confidentiality breaches inherent in third-party AI tools, and encouraging par- ties to include specific clauses on AI use in their Terms of Arbitration. These gaps carry legal consequences. Under Article 32, VIII, of Law No 9,307/1996, an arbitral award is null if rendered in violation of due process, equality of the parties or the arbitrator’s impartiality. If AI materially influences the tribunal’s deliberative process without proper disclosure, the resulting award may be vul- nerable to annulment. Arbitration in Brazil, moreover, does not operate in isolation from the judicial system. Foreign awards require recognition by the STJ; interim measures are sought before state courts; and annul- ment actions are processed by appellate tribunals. These proceedings are already subject to the same automated screening described above, meaning that the convergence between judicial automation and arbitral practice is a reality. Practical takeaways for corporate clients What is taking shape across Brazil’s dispute resolution landscape is a redefinition of the practitioner’s role. AI absorbs what is repetitive, voluminous and standard- isable, and in doing so it raises the baseline of what is expected from every participant in the system. For law firms and corporate clients navigating disputes in Brazil, the convergence of these trends yields several practical considerations. First, the automation of admissibility screening at the STJ means that the quality and precision of appel- late filings have never been more consequential. In a system where automated screening is, for practical purposes, dispositive, a poorly crafted special appeal is almost certain to be terminated without reaching a
77 CHAMBERS.COM
Powered by FlippingBook