The Verge reports that service workers across restaurants, parks, and schools are encountering a new kind of difficult customer: one who trusts AI chatbots over human expertise. A New York server described guests with shellfish allergies dismissing her warnings because ChatGPT said a dish was safe, and sommeliers being waved off in favor of chatbot wine picks, including bottles that do not exist.
The problem extends beyond dining. A river ranger said visitors arrive with AI-generated itineraries that include hallucinated campsites or impossible travel routes, and some remain skeptical even when shown a map. An after-school teacher said parents have used ChatGPT to set sleep schedules, ignoring staff advice that the child needed more rest.
The article also notes the rise of agentic AI: a restaurant manager received reservation requests and last-minute demands from AI agents on behalf of clients. The report suggests that as tools like Meta's Muse launch, workers will increasingly have to correct AI hallucinations while keeping customers satisfied.