Brazilian authorities said on November 13 that they are investigating a federal judge accused of using false information generated by the artificial intelligence (AI) chatbot ChatGPT to draft a ruling.
This is the first time such an incident has occurred in this South American country.
Specifically, the National Council of the Judiciary of Brazil (CNJ), the body that oversees the courts in Brazil, summoned Mr. Jefferson Rodrigues, a judge in the state of Acre, to explain why he published a judgment with many errors created by ChatGPT. Media reports said that in a trial, Mr. Rodrigues used many inaccurate information created by ChatGPT related to the judgments of higher courts in previous trials and applied these precedents to make a similar judgment.
In a document submitted to the watchdog, Mr. Rodrigues said the ruling was drafted by a “trusted advisor” with the help of AI. He explained that the use of AI to draft the ruling was due to the “excessive workload.” Mr. Rodrigues is expected to have 15 days to continue his explanation.
CNJ warns of a growing number of judges relying on AI, while Brazil has no legal regulations on this.
According to Tin Tuc newspaper