Across several US cities, diners have begun to notice that the glossy images on restaurant windows and menus are not photographs at all but creations of generative AI. A social-media post by user Maggie Moda, which has attracted millions of views, highlighted a series of oddly perfect sandwich and pastry pictures on New York sidewalks, prompting the question: what happened?
What happened?
Small eateries are using AI tools to generate photorealistic images of meals for promotional purposes. In San Francisco, the owners of Grind & Unwind, AJ Lozano and his partner Lyndsey Lozano, displayed AI-crafted dishes on a temporary menu while the restaurant was being set up. In Chicago, manager Ali Malik of Chi-Town Deli explained that similar images were used only until a new sign with real photographs could be installed. The images often feature perfectly layered croissant sandwiches, cheese that melts without a visible start, and burritos riddled with tiny holes that trigger trypophobia.
Why does it matter?
The practice revives a long-standing advertising trick: presenting food that looks better than it is. In 1965, the US Supreme Court considered a case where mashed potatoes were used to depict ice cream, establishing that an image is deceptive only if it misleads a reasonable consumer about the product's composition or quality. Legal scholars such as Rebecca Tushnet, a professor at Harvard Law School, note that the same principle applies today, an image that suggests a different size, ingredient list or appearance could be deemed misleading.
Advertising lawyers, including Libby O'Neill of Loeb & Loeb, warn that simply labelling an image as AI-generated does not automatically protect a business from claims of deception. The visual claim can convey information about quantity, ingredients or quality even without explicit wording, and regulators may still view it as a breach of consumer-protection rules.
What happens next?
Consumer backlash on platforms such as Reddit and TikTok suggests that shoppers prefer an honest, if imperfect, photograph of a real dish over a flawless but fictitious rendering. Industry observers expect that restaurants will either revert to traditional photography or adopt clearer disclosures about AI-generated content. Meanwhile, regulators in the United States and Europe are likely to revisit existing advertising guidelines to address the novel challenges posed by generative AI, potentially issuing new standards that require explicit labeling and verification of food images.
For now, the debate highlights a tension between cost-saving technology and the expectation of authenticity in food advertising. As AI tools become more accessible, the line between creative marketing and deceptive practice will continue to be tested.