The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Two lawyers in Pará, Brazil, were fined for using prompt injection—hidden instructions in legal documents—to manipulate the Galileu AI system used by the labor court. The concealed commands aimed to influence judicial decisions, undermining the integrity of the legal process. The court detected and penalized the misconduct.[AI generated]
Why's our monitor labelling this an incident or hazard?
The AI system of the Tribunal Regional do Trabalho was intentionally misled by the insertion of a hidden command designed to manipulate its output, which is a misuse of the AI system in a legal context. This manipulation led to a judicial response including fines and official condemnation, indicating that harm to the legal process and rights has occurred. Therefore, this qualifies as an AI Incident because the AI system's use was directly involved in causing harm related to legal rights and the justice system's integrity.[AI generated]
