The AI Slop Backlash: What It Means for Your F&B Tenant Mix

A few weeks ago, a post appeared in a Melbourne Malaysian food group I have been a member of for years, a community normally reserved for people trading tips on where to find proper char kway teow or arguing good-naturedly about the best roti canai outside of KL. Someone had posted a photo lifted from a local eatery's Facebook page: a plate of nasi lemak with a sambal that glowed an almost radioactive red, rice grains that repeated in an unnervingly perfect pattern, and a fried egg that looked like it was made out of plastic. Within an hour there were dozens of replies, ranging from gentle mockery to genuine vitriolic anger that a dish so tied to identity and memory for so many people in that group had been reduced to what one member called a cartoon of the real thing. A few said they would never order from that place, not because the food itself was bad, but because they felt genuinely insulted by the use of an AI generated image. I was taken aback by the intensity of the responses and subsequent posts from other establishments have elicited similarly angry responses. That Facebook thread was one small skirmish in a much bigger pattern. Restaurants, delis, and food trucks across Australia, Asia, and Europe have been quietly swapping real photography for AI-generated imagery, and diners everywhere are noticing, and pushing back, harder than anyone in the industry seems to have expected.

A Backlash Nobody Saw Coming

The examples keep stacking up, and they are global. Zomato, India's largest food delivery platform, moved to ban AI-generated food images back in 2024 after a spike in complaints from customers who felt misled about what they were ordering. In Malaysia, a Reddit post calling out a delivery listing for using obviously AI-generated photos sparked a similar pile-on, with commenters accusing the restaurant of being too lazy to take proper photos despite charging premium prices. Korean media outlets have documented the same pattern spreading through Japan, describing AI food photography that manages to look simultaneously overproduced and faintly alien, prompting one outlet to ask whether diners were now looking at alien food. Even here, threads compiling the worst AI menu offenders regularly circulate on Reddit and rack up tens of thousands of upvotes.

Food writers have started trying to explain why these images land so badly, including a recently published piece on CNN about appalling AI menu pictures going viral. An oft-cited piece of analysis from the food outlet Eater pointed to a kind of uncanny valley effect. We know, on some deep and specific level, what seared meat, cold noodles, or steaming rice are supposed to look like, and AI-generated versions tend to be just off enough to trigger discomfort rather than appetite, with repeating patterns, an unnatural gloss, or textures that do not quite appear the way food should present. The human brain notices even when we cannot quite articulate why.

What started as scattered social media mockery has hardened into a genuine consumer expectation. People want to see the food they are about to pay for, not a plausible facsimile of it. For cuisines that carry cultural weight, like the nasi lemak in that Melbourne group, the stakes feel higher still. Get the image wrong and it reads less like an unappetising photo and more like a lack of respect for the dish itself.

The Case for the Defence

Here is where the situation becomes complicated, because the operators reaching for these tools are very rarely trying to deceive anyone.

Most of the small businesses using AI-generated imagery are exactly that, small operations working with margins that do not stretch to a half-day shoot with a professional food photographer, let alone a graphic designer to lay out a proper menu. Think of a takeaway shop, a family-run deli, or a food truck. For years, these operators have had two choices. They could pay for imagery they could not really afford, or go without and rely on a laminated, text-only menu that does nothing to sell the dish. AI image generation looked, understandably, like a godsend. Suddenly a substantial photography budget became a few minutes and a subscription fee.

We have spoken with operators who genuinely believed they were levelling the playing field against bigger, better-resourced competitors. From where they stand, the criticism feels bewildering and unfair. Nobody is trying to trick anyone into ordering a dish that doesn't exist. They are trying to make an affordable banh mi look as good on a delivery app as a high-end plate at a place with an in-house creative budget.

Why It Lands Differently Anyway

The trouble is that intent and perception are two different things, and in hospitality, perception governs the business.

Diners are not evaluating AI-generated menu photography against the standard of being better than no photo at all. They are evaluating it against every other image they scroll past on Instagram, on delivery apps, or on a friend's story from dinner the night before. In that context, AI imagery does not read as resourceful. It reads as a shortcut, and worse, as a small, specific kind of dishonesty about the one thing a menu is supposed to promise.

There is also a broader mood shift at play that goes well beyond restaurants. AI slop has become a genuine shorthand across social platforms for content that is cheap, soulless, and slightly wrong. Once a category earns that label, anything that resembles it gets tarred by association, fairly or not. Authenticity has become one of the defining currencies of modern hospitality, whether it is provenance on a wine list, an open kitchen, or a chef's face on the door. For migrant and multicultural food communities in particular, visual accuracy is a matter of respect for provenance and cultural memory. An AI-generated hero shot cuts directly against that, even when the food behind it is entirely real, entirely handmade, and genuinely excellent.

The Fast-Food Fallacy: Why This Isn't Just Styling as Usual

It is tempting to dismiss the backlash as hypocritical. Hasn't the food industry spent decades living with commercial food styling, digital touch-ups, and glossy billboard burgers that look nothing like the grease-wrapped reality handed across the counter?

Equating traditional food styling with generative AI misses the point. Commercial styling enhances real ingredients to present an idealized version of an actual product. Generative AI manufactures a total fiction, often rendering physics-defying textures, bizarre geometries, and ingredients that do not exist in reality.

More importantly, consumers have a built-in, cynical contract with multinational fast-food giants where we expect corporate marketing machines to embellish. We go to local independents, neighborhood eateries, and cultural vendors for the exact opposite reason: human connection, proximity, and trust. When a local spot uses AI, it shatters that intimate contract. For culturally significant dishes, it stops looking like marketing polish and starts looking like a disconnect from the heritage of the food itself.

What This Means for Asset Managers

If you manage a portfolio of food and beverage tenants, whether that is a single retail strip, a shopping centre, or a mixed-use precinct, this is not a debate you can leave to individual operators to navigate alone. Brand trust, once dented, is expensive to rebuild, and a tenant's food imagery sits right at the point where a customer decides whether to walk in or keep scrolling. That makes it your reputational exposure too, regardless of lease clauses.

Precinct managers can build a few practical measures into their tenant oversight:

• Extend tenant audits past the fit-out. Most asset managers scrutinise shopfront design, signage, and lease compliance closely while ignoring what a tenant posts on delivery apps, social media, or menu boards. Add imagery and marketing collateral to regular health checks alongside the physical space.

• Understand why tenants reach for AI images. Most small operators using this technology are working with tight margins rather than trying to mislead anyone. Approaching this as a resourcing gap rather than a compliance failure yields better outcomes than a blanket ban buried in fine print.

• Consider shared imagery resources. A coordinated photography day, a modest imagery fund, or a preferred supplier list for affordable food photography can solve the underlying problem for tenants across a precinct at a fraction of the individual cost while removing brand risk.

• Fold imagery guidelines into onboarding. A short brand and marketing standard covering photography, menu language, and social content gives new tenants clarity from day one and establishes a consistent basis for discussions with existing ones.

  
  

None of this means your tenants should avoid AI tools altogether. Plenty of genuinely useful applications sit well away from customer-facing imagery, including rostering, inventory management, and concept ideation. The distinction worth holding onto is clear. Food is one of the last genuinely sensory purchases left, and customers can tell instinctively when what they are looking at did not come from a kitchen. Protecting your precinct's reputation means helping tenants find a practical way around that gap rather than leaving them to solve it poorly on their own.

Is your food and beverage tenant mix protected against reputational risk hiding in plain sight on the menu board? We work with developers and asset managers to audit brand experience across a precinct end to end, from fit-out to tenant imagery to the words on the page. Reach out to talk through where the gaps might be, or follow along here for more on the trends reshaping how we eat out.

#AISlop #FoodPhotography #RestaurantBranding #HospitalityTrends #FBStrategy #CustomerExperience #RetailStrategy