Dutch tech publication Tweakers has kicked off a legal storm after revealing that a Dutch lower court judge has used ChatGPT as a source of information for the ruling.
Tweakers says that this is the first time in Dutch case law that a judge has used ChatGPT.
The case was a neighbour dispute over solar panels, and the judge used ChatGPT to find out the average life span of a solar panel, as well as the average price of electricity. The judge did not use ChatGPT to help with the ruling.
The decision has been criticised locally by experts, including Henk van Ess, who is an expert in media literacy and fact checking, and says that everything that ChatGPT says should be taken with a grain of salt, telling one publication that he was “shocked by so much incompetence.” The issue is the lack of understanding of what ChatGPT is good for – which is not for fact checking. ChatGPT is still capable of hallucinating, aka giving entirely fictitious answers.
The Dutch judge is by no means the first to use ChatGPT in a ruling. UK Court of Appeal judge Lord Justice Birss in February last year used ChatGPT to provide a summary of an area of law, describing the chatbot as “jolly useful.” Sir Geoffrey Vos, Master of the Rolls, has cautioned about the need for ethical guidelines and proper regulation of the use of generative AI. However, the Courts and Tribunals Judiciary said in December that judicial office holders are personally responsible for material produced in their name and that provided they uphold confidentiality and privacy, and check accuracy, there is no reason why AI should not be used.
We’d argue that all judges ought to receive at the very least least some basic training on what generative AI is and does, and that until that happens, they should probably steer clear.
For an in-depth review of the latest and most up to date regulatory positions on the use of generative AI sign up HERE to our Gen AI Report out next week.
The image for this article was created by ChatGPT.