A central Victorian mayor and one-time whistleblower is launching authorized motion over what he alleges are defamatory statements made by ChatGPT.
Brian Hood, now mayor of Hepburn Shire, labored for Reserve Financial institution subsidiary Notice Printing Australia within the early 2000s, and alerted authorities that officers of NPA and one other subsidiary, Securency, have been paying bribes to win worldwide banknote printing contracts.
Hood’s attorneys Gordon Authorized defined that as a substitute of detailing Hood’s whistleblowing, ChatGPT “incorrectly recognized him as a person who confronted costs … somewhat than his precise function because the whistleblower within the case.”
The generative AI device makes a collection of false statements when quizzed about Hood’s function.
Gordon Authorized filed a ‘Considerations Discover’ to OpenAI on March 21 2023, detailing the inaccuracy and demanding a rectification.
“Brian Hood’s status as a morally upstanding whistleblower has been defamed by ChatGPT, which incorrectly represented him as a person answerable for illegally bribing international officers,” Gordon Authorized associate James Naughton mentioned in a press release.
“This crucial error is an eye-opening instance of the reputational hurt that may be attributable to AI techniques akin to ChatGPT, which has been proven on this case to provide inaccurate and unreliable solutions disguised as truth.
“As synthetic intelligence turns into more and more built-in into our society, the accuracy of the knowledge offered by these companies will come beneath shut authorized scrutiny.
“The declare introduced will intention to treatment the hurt induced to Mr. Hood and make sure the accuracy of this software program in his case.”
