What Happened
On 21 October 2026, Just Eat issued a public apology after a wave of social‑media ridicule over AI‑generated food images that appeared on its platform. The company had been testing a new AI system to create pictures for menu items that lacked photographs. Instead of realistic depictions, the model produced bizarre concoctions: bacon and eggs floating in a glass of cola, chips poking out of muffins, and a “cheese roll” that looked nothing like its real‑world counterpart. A Just Eat spokesperson said the firm was “in the process of removing” the incorrect images and would halt further AI‑generated pictures until the technology was refined. The spokesperson added, “We work with tens of thousands of small restaurants on their food imagery, including funding professional local photographers. We’ve been trialling AI‑enhanced images, and as with any new technology, it isn’t always quite right the first time.”
What This Means For You
If you’re a small‑restaurant owner using Just Eat to showcase your menu, this incident highlights the importance of vetting AI‑generated content before it reaches customers. Even when an AI claims to “enhance” images, the output can misrepresent your food, potentially damaging brand trust. Consider setting up a quick review step: have a staff member or a freelance designer approve any AI‑created images before they go live. This extra layer of quality control can prevent embarrassing mistakes and keep your menu accurate.
For delivery‑app users, the episode serves as a reminder that AI can hallucinate. When you see a picture labeled “AI‑generated,” treat it with a healthy dose of skepticism. If the image looks off, double‑check the menu description or look for a photo from the restaurant’s own social media. This practice can help you avoid ordering a dish that looks nothing like the real thing.
If you’re a developer or product manager working on AI‑driven content generation, the Just Eat case underscores the need for robust validation pipelines. Implement automated checks that flag improbable combinations—such as a beverage containing solid food items—or use human‑in‑the‑loop reviews for high‑impact outputs. Also, maintain clear labeling so users know when an image is AI‑generated versus a genuine photograph.
Finally, for anyone interested in AI ethics, this incident is a textbook example of how hallucinations can erode public trust. Companies deploying generative models should communicate transparently about limitations and actively involve stakeholders—customers, creators, and regulators—in shaping deployment strategies.
Why It Matters
This episode illustrates a broader trend: generative AI is still prone to hallucinations that can have real‑world consequences. When a food‑delivery platform misrepresents menu items, it can lead to customer disappointment, negative reviews, and reputational damage. The incident also shows that even well‑intentioned AI experiments can backfire if not carefully monitored.
From a regulatory perspective, the fallout could prompt tighter scrutiny of AI‑generated content in commercial settings. If consumers feel misled, consumer protection agencies might investigate whether companies are providing accurate representations. This could lead to new guidelines or standards for labeling AI‑generated media, similar to the EU’s proposed AI Act.
For the AI industry, Just Eat’s experience is a cautionary tale that emphasizes the need for better alignment between model outputs and real‑world expectations. It also highlights the value of human oversight, especially when the stakes involve customer perception and brand integrity.
Key Takeaway
- Just Eat’s AI‑generated images caused public ridicule, prompting an apology and a pause on further AI content.
- Small restaurants should review AI images before publishing to avoid misrepresentation.
- Users should verify AI‑labeled images against actual product descriptions or photos.
- Developers must implement validation pipelines and clear labeling to maintain trust.
Frequently Asked Questions
What caused the AI to produce such odd images?
The AI was trained to generate food photos from textual descriptions, but it misinterpreted terms like “soda” versus “soda bread,” leading to unrealistic combinations such as bacon and eggs in cola.
Will Just Eat resume using AI for images?
Yes, the company said it will stop generating new AI images until the technology is further developed and reliable.
How can I tell if a menu image is AI‑generated?
Just Eat labels AI‑generated images on the app. Look for a small tag or note indicating the image was created by AI.
Leave a Reply