‎Brazil News

 

 

The National Council of Justice (CNJ) will investigate a federal judge from the Federal Regional Court of the 1st Region (TRF1) for using the ChatGPT generative artificial intelligence tool in preparing a judicial decision.

 

The lawyer, representing the defeated party in the action, identified the use of the tool because of non-existent jurisprudence in the sentence, and formalized a complaint with the Regional Federal Justice Inspectorate of the 1st Region. The investigation into the case had been archived by the Internal Affairs Department, but was reopened by the CNJ.

 

In his defense, the judge stated that the sentence had been made by a server from his office and attributed the use of ChatGPT to a “mere mistake”, resulting from the “work overload that falls on the shoulders of the judges” at TRF1.

 

 

In a statement, the regional inspector of the Federal Court of the 1st Region, judge Néviton Guedes, stated that it came to his attention that the tool was used “as an assistant in drafting a judicial act and presented non-existent precedents in the results of the jurisprudential research”.

 

Guedes recommended not using generative artificial intelligence tools not approved by the control bodies of the Judiciary in researching jurisprudential precedents.

 

The judge highlighted the need to carry out studies on new technologies to improve judicial activity and reinforced the duties of caution, supervision and responsible disclosure of process data, regarding the use of AI in preparing decisions.