DistantNews
Support us
AI in the Judicial Process: Cognition Versus Algorithms
🇧🇷 Brazil /Technology

AI in the Judicial Process: Cognition Versus Algorithms

From Estadão · () Portuguese

Translated from Portuguese, summarized and contextualized by DistantNews.

At a glance

News Named sources Context piece
  • Two lawyers in Brazil's Pará state are accused of attempting to bypass an AI system in a labor court by inserting hidden commands into their legal filings.
  • The technique, known as 'prompt injection,' aims to manipulate AI behavior, in this case, to prevent a thorough analysis of evidence.
  • This incident raises ethical questions about linguistic manipulation in legal proceedings and the future of human interaction within the judicial system, especially with the increasing use of generative AI.

A recent incident involving two lawyers in the state of Pará has brought to light the complex ethical and practical challenges posed by the integration of Artificial Intelligence into Brazil's judicial system. The lawyers are accused of attempting to subvert the 'Galileu' AI system of the Regional Labor Court of the 8th Region by embedding imperceptible commands within their legal documents. This maneuver, termed 'prompt injection,' is designed to manipulate the AI's analysis, specifically to avoid a critical examination of the evidence presented in a judicial case.

While the lawyers' actions are described as excessive and reprehensible, the case prompts a deeper reflection on the boundaries of persuasive legal argumentation and the use of linguistic artifices within the judiciary. Traditionally, legal processes have been inherently human-driven, relying on the discursive construction of arguments and the subjective interpretation of judges, influenced by their own beliefs and worldviews. This human element, with its inherent nuances, fragilities, and cognitive biases, has historically shaped the 'truths' within legal proceedings.

The introduction of Generative AI, however, represents a significant shift. It's not merely another technological tool but a new algorithmic logic capable of profoundly altering work dynamics and repositioning the judicial process. This development challenges the long-standing 'equality of arms' that characterized legal disputes, which operated within the confines of human cognition. As psychologist Daniel Kahneman's work illustrates, human decision-making is not strictly linear or rational; it involves cognitive biases and heuristics. Yet, these limitations were, to some extent, shared among all participants.

This case, as reported by Estadão, highlights a critical juncture for the Brazilian judiciary. It forces a re-evaluation of how AI should be employed, ensuring it enhances justice rather than becoming a tool for manipulation or undermining the fundamental guarantees of due process and ample defense. The incident serves as a stark reminder that as AI becomes more embedded in legal practice, the ethical considerations surrounding its use, and the potential for its misuse, must be at the forefront of discussions.

DistantNews Editorial

Originally published by Estadão in Portuguese. Translated, summarized, and contextualized by our editorial team with added local perspective. Read our editorial standards.