Workers say customers are using ChatGPT to dispute basic facts

In short: More customers are treating AI chatbots like ChatGPT as an authority, even when the chatbot is wrong, and workers say it is making everyday interactions harder.

What's going on

A report from The Verge describes a growing pattern in restaurants and other service jobs. Customers are using generative AI tools, like ChatGPT, to make decisions and to argue with staff.

One example comes from Madison, a server in New York City. She says she asks diners about allergies at the start of each meal. Recently, she has seen “close calls” where guests say they have a shellfish allergy, then order fish dishes that include shellfish-based broth.

Madison says that when she warns them about ingredients, some guests push back by saying ChatGPT disagrees. This is tied to a known problem with chatbots called “hallucinations,” which is when an AI confidently makes something up (like a friend who answers even when they do not know). In these cases, the wrong answer can create real risk, since food allergies can be serious.

Madison also says guests are increasingly using ChatGPT instead of asking trained staff for help, like a sommelier who knows wine. She says she has suggested bringing a sommelier to the table, and guests reply that they will “just ask ChatGPT.”

What to watch

This trend may force more businesses to set clearer rules about when AI answers are acceptable and when staff instructions come first, especially for safety issues like allergens. It also raises a bigger question, who is responsible when an AI gives bad advice and a customer acts on it.

Source: The Verge AI