There’s a growing ethical debate around the world about how people use technology to create and distribute information. UNESCO highlights cases where AI raises questions about responsibility, transparency, human oversight, and potential harm. These concerns make sense especially when AI-generated images present something as ordinary as a restaurant meal.
The main issue here is that this technology has moved far beyond its original niche. A UK Government survey found that 79% of adults had used at least one AI tool, while 34% had used generative AI tools. Those figures show how quickly the tech has entered everyday life, from work tasks to creating images and other content.
That rapid adoption could soon make AI-generated material difficult to distinguish from human-created content online. The prediction is that 90% of online content could become AI-generated by 2026. If this comes true, people will encounter synthetic images and text constantly, including when they browse restaurant menus, websites, delivery apps, and social media.
Ever heard of the "Dead Internet Theory"? It claims that much of what we see online no longer comes from real people, but from bots and AI-generated content. The theory emerged on fringe forums in the late 2010s and argued that the internet is gone forever as automated content began taking over search results and social feeds.
While the theory’s literal claim remains speculative, the European Consortium for Political Research notes that its central idea has become harder to dismiss: the digital world increasingly contains automated content that looks human. The issue is that machine generated content is moving from the digital to the real world.
However, AI has also created new jobs and roles. Businesses have built entire workflows around the technology, spawning more than 1.3 million new AI-related jobs. AI engineers, data annotators, AI trainers, ethics specialists, all these show how quickly the industry is adapting.
Some of those emerging roles exist because businesses need people who can understand artificial intelligence. Meanwhile, companies increasingly rely on AI for marketing and visual content, creating another everyday use for these tools.
At the same time, AI can do more than create an image of a meal. Researchers have also used generative AI to design actual burger recipes, training a model on 2,216 human-created recipes before generating and testing new combinations. Some of those burgers performed as well as or better than a Big Mac in blind sensory tests.
The goal was also to develop burgers with specific nutritional and environmental goals. A mushroom-based recipe achieved a much lower environmental impact, while a bean-based burger scored substantially higher for nutrition. In this case, AI did not simply make food look appealing, but it helped create food that humans could actually taste and evaluate.
Personally, what I find really weird about these food ads is that they seem completely deceitful. The European Union rules require food information to avoid misleading consumers, including information about a food’s characteristics, composition, quantity, and presentation.
A restaurant could face an even stricter situation in Japan, where food labeling and presentation rules place significant emphasis on accurate information about products. An AI image that dramatically changes what a customer should expect would raise serious compliance concerns.
The situation looks way less specific in the United States, where most of the images posted online come from. The Federal Trade Commission suggests truth-in-advertising principles, but there doesn’t appear to be a federal rule that simply says a restaurant’s advertised food pic must exactly match the meal served to customers.
