A judge in Colombia enlisted the help of the AI text generator ChatGPT to formulate a verdict. Juan Manuel Padilla of the Labor Court in the port city of Cartagena used the tool to argue why an autistic child’s insurance should pay for medical expenses, Bogotá-based BluRadio reports.
The radio station also published the questions to the AI, formulated in Spanish, and their answers. Accordingly, she has affirmed, for example, that minors with autism are exempt from paying therapy costs and named the associated legal text. The tool has thus taken over a job that a secretary would otherwise do, Padilla said accordingly.
“Don’t make judges lazy”
Judges not only have to know about developments in the judiciary, but also about technology, Padilla explained in an interview with BluRadio. An AI like ChatGPT helps to create “very understandable sentences with good formulations”. In addition, such a tool could help establish equality by referencing precedents. Recourse to such a tool does not make a judge lazy, his decision remains autonomous and his own. He understands the concerns associated with resorting to ChatGPT, but knows a group of judges who are very happy with the approach. You would know that this technique can help to transcribe stuff from the internet and summarize it in a logical and concise way.
Meanwhile, the British Guardian points out that there has also been criticism of Padilla’s initiative. With Juan David Gutierrez, a professor at Del Rosario University was “stunned” and called for urgent training in “digital competence” for judges. At the same time, the British newspaper points out that just last year a law was passed in Colombia that requires lawyers to use technology to make their work more efficient. A Supreme Court judge has prophesied that the use of ChatGPT in courts should soon be accepted and widespread.
But the chatbot itself has its doubts. In response to a request from the Guardian, ChatGPT has advised judges not to use the tool for judgments. It is not a substitute for a person’s “knowledge, expertise and judgment”. In view of the hype about the new technology, it has been pointed out several times that ChatGPT repeatedly formulates false statements with a self-confidence that conceals the errors.
(my)
To home page
