One of the biggest concerns regarding ChatGPT’s responses is reliability. OpenAI may be working on making the generative AI chatbot more trustworthy, but that doesn’t change the fact that chatbots can and will provide incorrect data. The best example of that is Google’s big Bard fumble, which tanked the company’s stock a few weeks ago.
ChatGPT might reach a new milestone, becoming the first generative AI product to be the target of a lawsuit. A mayor in Australia is considering legal action against OpenAI and ChatGPT, as the chatbot shared false claims about him. This would be the first defamation suit against generative AI and yet another legal problem OpenAI would have to deal with.
A few days ago, an Italian privacy watchdog banned ChatGPT, citing concerns about user data handling. The regulator also mentioned the chatbot’s inaccuracy in its announcement. Other countries might take similar action.
Other public figures concerned about their reputations might follow Brian Hood’s lead. The man became the mayor of Hepburn Shire in Australia last November. He’s now worried about ChatGPT, as members of the public told him that the generative AI product named him as a guilty party in a bribery scandal going back to the early 2000s. The matter involved a subsidiary of the Reserve Bank of Australia, Note Printing Australia.
Lawyers representing Hood told Reuters that Hood did work for the company. But he was the person who notified authorities about bribes paid to foreign officials to win printing contracts.
Hood’s lawyers sent a letter of concern to OpenAI on March 21st. The company has 28 days to fix the errors or face a defamation suit.
It’s unclear how fast OpenAI would be able to correct the issues. And whether ChatGPT will stop making errors concerning this particular public figure. Also, Hood’s case might inspire others to take similar action if they think the chatbot is spreading false claims.
Should Hood go forward with a trial, “it would potentially be a landmark moment in the sense that it’s applying this defamation law to a new area of artificial intelligence and publication in the IT space.” That’s what James Naughton, a partner at Hood’s law firm Gordon Legal, told Reuters.
The report explains that Australian defamation damages payouts top around A$400,000, or nearly $270,000. Hood doesn’t know how many people had access to ChatGPT’s false information about him. That’s a factor in determining the size of the payout. But Naughton believes Hood is entitled to at least half of that top payout given the seriousness of the defamatory claims.