OpenAI, the company behind popular AI chatbot ChatGPT, is facing a libel lawsuit after wrongly accusing a man of embezzlement. The incident occurred after a journalist asked ChatGPT to provide a summary of a legal case involving Mark Walters, a radio host who was accused of embezzling funds from the Second Amendment Foundation. ChatGPT provided false information, claiming that Walters had worked for the foundation when, in fact, he had not. Walters is now suing OpenAI for libel, with the case filed in Gwinnett County Superior Court in June.
This is not the first time that ChatGPT has caused problems for OpenAI. Last month, the generative AI model produced fake legal citations, causing a lawyer to use fabricated cases in his legal filings. Researchers have identified the issue as hallucination, where AI models fabricate information and present it as fact. Speaking at IIIT Delhi, OpenAI founder and CEO Sam Altman admitted that he trusts the answers that come out of ChatGPT the least out of anyone on earth.
The case highlights the challenges associated with developing advanced AI models that can balance creativity and accuracy. While countries around the world have introduced national AI strategies, the industry still needs to tackle issues such as hallucination and ensure that AI-generated output is more reliable. For now, OpenAI will focus on perfecting the ChatGPT model to reduce such incidents in the future.
