A New York City server reports that diners are using ChatGPT to challenge her safety warnings regarding food allergens. Madison, who works as a server in New York City and requested anonymity to protect her identity, describes a pattern of close calls at her tables. She begins service by asking every guest about allergies, a standard practice in the industry. However, she has encountered situations where a diner discloses a shellfish allergy, yet she returns to the table to find no follow-up questions from the guest. In these instances, the diner proceeds to order a fish dish that is prepared with a shellfish-based broth. When Madison informs the customer that the dish contains their allergen, the guest pushes back. They assert that ChatGPT disagrees with her assessment, stating that the dish does not contain shellfish. Madison explains that these customers prioritize the AI output over her direct warning. "They just want to talk to ChatGPT," she says. The incident highlights a growing trend where individuals rely on generative AI for factual verification in real-time scenarios, even when that information contradicts professional advice on health and safety.