Mayor of Hepburn Shire, Australia, Brian Hood, said he may sue OpenAI, owner of the artificial intelligence (AI) ChatGPT, amid reports that the chatbot has made false claims that he served time in prison for bribery.
ChatGPT has reportedly said that Hood went to prison in connection with a foreign bribery scandal involving a subsidiary of the Reserve Bank of Australia, Note Printing Australia, in the early 2000s. His lawyers said he did work there but that he was the one who notified authorities of the scheme.
While the slandering of Hood's reputation is troublesome, it will be very difficult to prove an AI algorithm is at fault for disseminating defamatory information. To defame someone — legally speaking — the perpetrator must have knowingly disseminated the falsities with malice, but how could a computer do that? Such cases involving public figures will lead to ever-growing debate on the issue of AI and its role in public discourse, but to sue a robot isn't a winnable course of action.
The case against OpenAI has nothing to do with the algorithm and everything to do with the company's delayed response to Hood's request. Once Hood proved the information to be false, OpenAI should have scraped it off the platform immediately, but, according to the lawyers, it hasn't done that and therefore has opened itself up to a legitimate allegation of defamation.