BRAZIL Trends and Developments Contributed by: Eduardo Perazza, Glaucia Coelho, Débora Chaves Martines Fernandes and Agnes Domingues, Machado Meyer
Machado Meyer, Sendacz & Opice Rua José Gonçalves de Oliveira No 116 5º andar Itaim Bibi São Paulo, SP Brazil 01453-050 Tel: +55 11 3150 7000 Email: bps.mkt@machadomeyer.com.br Web: www.machadomeyer.com.br/en
Artificial Intelligence and Dispute Resolution in Brazil: Where Algorithmic Efficiency Meets Strategic Advocacy Brazil’s dispute resolution landscape is being reshaped by a tension that jurisdictions worldwide are grappling with but none has yet resolved: the grow- ing reliance on artificial intelligence (AI) to manage an unmanageable volume of disputes, and the simultane- ous acknowledgement that the most consequential cases – whether before the superior courts or in high- stakes arbitrations – still demand the kind of strate - gic, human judgment that no algorithm can replicate. Nowhere is this tension more visible than at the top of the Brazilian judiciary. The Supreme Court (STF) exercises constitutional review; the Superior Court of Justice (STJ) standardises the interpretation of federal legislation. Neither was designed as the entry point of the justice system, but as its selective apex, open only to genuinely novel legal questions, interpretative divergencies, binding precedents or issues of consti- tutional relevance, where facts are not discussed and evidence is not re-examined. It is this paradigmatic function that makes access inherently restrictive, and that makes so stark the contradiction between the vol- ume that actually reaches these courts – hundreds of thousands of cases per year – and the human capac- ity to process it. It is here that AI has found fertile ground to consolidate itself as a central component of judicial activity. According to the STJ Presidency Report (2024–2025), the Court’s 33 justices received nearly 500,000 cases to decide in 2025, a caseload without parallel among consolidated democracies. The scale of the challenge
is best appreciated through comparison. Brazil has roughly 215 million inhabitants. In France, the Court of Cassation and the Council of State – two apex courts with distinct jurisdictions – together count approxi- mately 200 members and decide between 10,000 and 20,000 cases per year each. In India, with 1.4 billion people and more than six times Brazil’s population, the Supreme Court, composed of 34 justices, issues around 1,000 reported judgments annually; a frac- tion of the volume that lands on the desks of STJ’s justices. These comparisons are necessarily imper- fect – the institutional design, procedural rules and filtering mechanisms vary widely between jurisdic- tions – but they serve to illustrate the extraordinary dimension of Brazil’s caseload problem, especially in the upper levels of the judiciary. Faced with this real- ity, STJ invested USD52 million (BRL297.4 million) in new technology solutions in 2025 and placed gen- erative AI at the centre of its response to congestion. Today, before any justice becomes aware of a case, AI has already played a role in the preliminary filtering of admissibility, a development that is reshaping the way litigation before the superior courts is conducted. This article traces the arc of that transformation across Brazil’s dispute resolution landscape. It begins with the superior courts, where AI has moved from classify- ing cases to drafting decisions, and where automated admissibility screening has fundamentally altered the conditions of access to the highest judicial instances. It then broadens the lens to the legal profession and to international arbitration, moving to examine how practitioners, arbitral institutions and corporate clients are navigating an environment in which standardised
73 CHAMBERS.COM
Powered by FlippingBook