The slowness of Brazilian Justice is a historical bottleneck, a challenge that has dragged on for decades, impacting millions of citizens and businesses. With one of the largest volumes of lawsuits in the world, the system operates under constant overload, resulting in proceedings that extend for years. However, a seismic shift is underway. The National Council of Justice (CNJ) has just nationalized “Berna” AI, a tool developed by TJGO that promises to be the catalyst for a revolution in procedural efficiency, tackling the most acute pain point: the sluggishness in the initial phase of lawsuits.
What Happened
“Berna” (Recursive Electronic Search Using Natural Language) is an artificial intelligence created by the Goiás State Court of Justice (TJGO) to automate the triage and initial processing of lawsuits. Its capability lies in reading and interpreting judicial petitions, automatically classifying the type of action, the requests, and the parties involved. This allows for faster and more standardized organization and distribution of cases, suggesting movements and internal workflow paths that previously consumed hours of human labor. The CNJ’s strategic decision to nationalize Berna means that this AI will be made available to all courts in Brazil – state, federal, and labor. This is not a simple replication; the CNJ will standardize procedural classification, integrate Berna into the Electronic Judicial Process (PJe), and establish minimum parameters for transparency, auditing, and data security. In practical terms, we are talking about the automation of highly repetitive initial tasks, freeing up civil servants and advisors to focus on activities that require human discernment and legal complexity, directly addressing the entry point of the procedural funnel where scale gains are massive.
The Alchemist’s Analysis
The movement to nationalize Berna, although focused on a single artificial intelligence, is a precursor to what Centrato AI has advocated as the era of “multi-agent intelligence systems.” Think about it: Berna, in its essence, acts as an agent specialized in triage. However, the true quantum leap in efficiency and robustness does not lie in a single centralized algorithm, but in the orchestration of multiple AIs, each optimized for a specific function. A risk analysis agent, another for precedent summarization, a third for identifying argumentative gaps, and so on. An isolated system is effective, but an ecosystem of interoperable agents is transformative. Berna, by integrating with the PJe and operating on a national scale, establishes the infrastructural and cultural foundations for such “multi-agents” to flourish. To ignore this progression is to underestimate the disruptive potential of AI. A single agent, however sophisticated, is just the beginning; the future belongs to the coordinated synergy of distributed intelligences, which will work together to break down complexity and accelerate decision-making in all spheres.
Operational Impact
Berna’s impact transcends mere task automation. Its nationalization redefines the Judiciary’s operational structure under three critical pillars: security, governance, and orchestration.
- Security: The handling of sensitive data in millions of petitions raises rigorous protection imperatives. AI must ensure the confidentiality and integrity of information, preventing leaks and misuse. Algorithmic security is equally vital: how to mitigate historical biases in training data that could reproduce or amplify inequalities in the decision-making process? Responsibility for systemic errors also needs to be clearly defined.
- Governance: The adoption of AI in such a sensitive sphere demands a robust framework. This includes transparency of classification criteria, algorithmic auditability, and compliance with Brazil’s emerging legal framework for AI (Bill 2.338/2023). Who monitors Berna’s performance? How is its impartiality and explainability ensured? These are fundamental questions for legitimacy and public trust.
- Orchestration: Berna does not operate in isolation. Its integration with the PJe and the re-engineering of judicial workflows are crucial. Automation frees up time but requires civil servants and magistrates to adapt their roles, shifting from repetitive tasks to supervision, validation, and exception management. The optimization of human and technological resources requires intelligent orchestration, ensuring that AI amplifies, rather than replaces, human judgment and analytical capacity.
Conclusion
The nationalization of Berna AI by the CNJ is more than a mere technological update; it is a watershed moment that accelerates the modernization of the Judiciary and sets a precedent for the integration of artificial intelligence into large-scale public services. However, the success of this initiative lies not only in Berna’s technical capabilities but in how we address the governance, security, and ethical challenges that AI inevitably brings. It is an invitation for deep reflection on the future of legal work, the redefinition of roles, and the construction of a more agile and equitable justice system. This is the essence of pragmatic intelligence: embracing innovation without losing sight of the fundamentals of responsibility and human impact. To delve deeper into how artificial intelligence is shaping the future of strategic operations and to not miss the Alchemist’s next analyses, subscribe to our newsletter and explore our exclusive AI implementation methodology.
