A man in Singapore who was applying for a Personal Protection Order (PPO) was found to have used ChatGPT to research for his written submissions in court.
He ended up citing 14 fictitious cases, according to a judgment made available on Wednesday (10 Sept) and seen by Channel NewsAsia (CNA).
Source: SG Courts
Both the man and his ex-wife had been seeking PPOs against one another for themselves and their two children after their divorce was finalised.
A PPO is a court order restraining the respondent from committing family violence against the applicant or other family members.
Those who breach a PPO may be jailed and/or fined.
Family court magistrate Soh Kian Peng was considering the PPO applications from both former spouses, with the man representing himself in the case.
Source: Kameleon007 on Canva
However, the magistrate realised that none of the 14 cases cited by the man existed.
He described some of them as “quite obviously fictitious”, while others that appeared to be legitimate at first were not.
The man had said they were relevant local legal examples that were similar to his case, and could thus be used as legal precedents.
However, he admitted that he had used ChatGPT to help identify the cases and failed to verify them before citing them in court documents.
The man also referenced provisions of the Women’s Charter that were outdated and no longer applicable.
Mr Soh noted that court users were responsible for all the content in their court documents.
That means they need to ensure that any output generated by artificial intelligence (AI) is accurate and relevant before submitting it.
The man had said that he was unaware of this guideline.
Source: Emiliano Vittoriosi on Unsplash
The consequences for not checking AI-generated output included the submission being disregarded and having to pay the other party’s legal costs.
Eventually, the court ultimately dismissed both PPO applications, ruling that neither parent was likely to commit family violence on their children.
However, the ex-husband was ordered to pay costs amounting to S$1,000 to his ex-wife.
Mr Soh also told the man to declare in writing if he uses generative AI to prepare any court-submitted documents in future.
He would also be expected to state that he had complied with the relevant guide, to guard against him citing “AI-hallucinated material to the court”.
The magistrate clarified that this did not prohibit the man from using generative AI, but would prevent the court from being “taken by surprise”.
Also read: Man uses AI as lawyer in New York court, judge not amused & turns it off
Have news you must share? Get in touch with us via email at news@mustsharenews.com.
Featured image adapted from SG Courts and Emiliano Vittoriosi on Unsplash.