AI can still get things wrong. OpenAI's own researchers say language models sometimes guess when they are uncertain, and produce plausible but incorrect statements. Good setup reduces this a lot, but it does not remove it.
You are also responsible for what it says. In Moffatt v Air Canada, decided in February 2024, a British Columbia tribunal found the airline liable after its chatbot wrongly told a customer he could claim a bereavement discount after travelling. The tribunal said Air Canada could not separate itself from the chatbot on its own website. It awarded the customer around 650 Canadian dollars in damages, plus interest and fees. It is a Canadian decision, not UK law, but it shows how a tribunal may see it.
Chatbots can also be pushed off track. In January 2024 the parcel firm DPD switched off the AI part of its chat after a customer got it to swear and criticise the company. DPD said the problem came after a system update. Any assistant open to the public needs limits on what it will talk about.
Treat every answer your assistant gives as if you had written it on your website yourself. If you wouldn't publish it, don't let the assistant say it.