AI hallucinations: fact vs fiction

• While we await specifics on AI regulation in the UK, this article examines the libel risk in this jurisdiction faced by users of generative AI systems who republish false output data.
• It examines the importance of companies devising internal policies on the use of generative AI in the workplace which properly factor in the risks faced by users republishing false output data about third parties.

It is well-publicised that ChatGPT recently invented a sexual harassment scandal, naming a real law professor as the accused (citing a fake Washington Post article as evidence in support of the allegation). Not only did no such article exist, but the real professor had never been accused of harassing a student, nor had he been present on the trip to Alaska described by the chatbot during which the