AI is changing how shoppers decide what to eat, what to buy and which brands deserve a place in the basket. These findings are based on a UK consumer survey conducted in May 2026, where shoppers were asked to use AI tools to ask questions about healthy eating and share the answers they received. The final dataset included 1,286 AI-generated responses from UK shoppers across tools including ChatGPT, Gemini, Copilot, Claude and Amazon Rufus.
The analysis shows how AI currently interprets healthy eating questions, which themes it returns to most often, and where brands have a chance to become more visible, useful and recommendable in AI-generated advice.
This is not theory about a distant shopping habit. The questions were practical. Shoppers asked whether foods were healthy. They asked what to buy for weight management. They asked for high-protein snacks, lower-sugar treats, lunchbox ideas, healthy swaps and meal plans.
AI did not answer like a supermarket shelf. It did not show every brand. It gave selective advice. Sometimes it named products. More often, it recommended types of products, retailers, ingredients, formats and rules of thumb.
That matters. If AI cannot clearly understand what job a brand does, it is unlikely to mention it.
The Top 10 Actions Brands Can Take:
For brands, the work is not to shout “healthy” more loudly. The work is to make the product easy for AI to place in a real eating situation.
“Healthy” is too broad to be useful.
AI needs sharper signals. A brand should be easy to connect with a specific consumer job, such as:
The largest question type was specific food healthiness and best food questions, at 12.0% of cleaned questions. Shoppers repeatedly asked AI whether a food was healthy, what the healthiest option was, or what they should choose for a particular need.
That creates an obvious gap. A brand that only says it is “better for you” gives AI very little to work with. A brand that says it is a high fibre afternoon snack, a lower sugar evening treat or a plant-based lunchbox option is much easier to place in an answer.
A weak AI answer says:
Choose a healthy snack.
A stronger answer says:
For a high fibre afternoon snack, look for portioned products that provide fibre, taste and convenience.
Best fit supermarket categories:
How to do it
Write the product’s healthy eating job in plain English and repeat it consistently across the brand site, retailer pages, FAQs and product descriptions.
AI is careful with nutrition claims. That is a good thing. It also means brands need to give it language it can safely repeat.
Useful claims include:
Nutrients, macros and functional ingredients were the second largest question type, at 10.4%. Shoppers asked about protein, fibre, calories, vitamins, minerals, energy, fullness and healthy ingredients.
Across responses, AI tended to value the same things: protein, fibre, whole foods, lower sugar, lower salt and balance.
That does not mean every brand should chase the same claim. It means each brand should make its strongest real claim easy to read, easy to verify and easy to repeat.
A vague brand page forces AI to fall back on generic advice: Eat more protein and fibre.
A clearer page helps AI say: Look for products clearly labelled as high in protein or a source of fibre.
Best fit supermarket categories:
How to do it
Strip out loose health language. Replace it with accurate claims, written in the same form across every main product surface.
Fresh reviews and FAQs help AI understand what the product is, why people use it, what need it serves and how shoppers talk about it.
Your website should include:
Across the data, AI responses were more likely to mention broad categories than specific packaged food brands. Retailers were also more visible than brands.
That suggests many brands are not yet distinct enough, trusted enough or easy enough to retrieve in AI healthy eating advice.
Fresh reviews help AI understand consumer language around taste, texture, convenience, satisfaction and usage occasions. FAQs help AI answer the questions shoppers are already asking.
The aim is to move from: Choose a healthy snack.
To: For this need, this product is a credible option because it has the right claim, format, usage occasion, consumer feedback and availability.
How to do it
Treat reviews and FAQs as part of the product’s AI visibility, not as an afterthought for the website team.
Most product pages still read as if shoppers arrive knowing what they want. AI questions show the opposite. Shoppers often arrive with a problem.
Useful pages and FAQs could answer questions such as:
The survey showed shoppers using AI for practical decisions. Shopping lists, supermarket choices and product recommendations made up 8.6% of cleaned questions. Healthy recipes and meal ideas made up 5.8%. Meal plans and structured eating plans made up 4.7%. Budget friendly healthy eating made up 4.5%.
These are shopping questions hiding inside advice questions.
A brand that only describes the product misses that moment. A brand that answers the actual prompt has a better chance of being picked up and used.
Best fit supermarket categories:
How to do it
Build FAQs from real shopper language. Use the same questions people ask AI, then answer them clearly and briefly.
AI already uses swap logic. Brands should use it too.
Shoppers asked for alternatives to crisps, chocolate, sugary cereals, fizzy drinks and skipped meals. That links directly to specific food healthiness at 12.0%, shopping and product recommendations at 8.6%, and weight management at 9.6%.
Examples:
This is where many brands are too polite. They describe the product but avoid saying what it replaces. AI needs the comparison.
A general answer says: Choose healthier snacks.
A more useful answer says: If you want something crunchy instead of crisps, look for popcorn, roasted chickpeas or portioned savoury snacks.
Best fit supermarket categories:
How to do it
Create swap pages and retailer content that name the old habit and the better alternative.
New products should be built against needs AI is already seeing in consumer questions.
The most visible needs in the data were:
The product cues that appear most useful include:
A new product should answer a clear question.
“Better for you” is not a launch platform. It is a holding phrase. AI visible needs are sharper and more commercial.
Best fit supermarket categories:
How to do it
Test innovation ideas against AI prompts before launch. If the product does not answer a question shoppers are asking, sharpen the proposition.
Consumers often asked AI for structure. Healthy recipes and meal ideas made up 5.8% of cleaned questions, while meal plans and structured eating plans made up 4.7%.
Brands should stop assuming the occasion is obvious.
Useful examples include:
This is simple work, but it is often missing. AI can recommend a product more naturally when it can explain where it fits.
Best fit supermarket categories:
How to do it
Add meal, snack and usage occasions to product pages, FAQs, recipe content and retailer copy.
Retailers were much more visible than individual food and drink brands in AI responses.
Aldi, Tesco, Lidl, Sainsbury’s and Asda appeared far more often than packaged food brands. That links directly to shopping lists, supermarket choices and product recommendations, which accounted for 8.6% of cleaned questions.
AI often answers healthy eating questions through access and affordability. It wants to know where the product sits in a real shop.
Brands should make it easy to understand:
A weak answer says: Buy healthy snacks at the supermarket.
A stronger answer says: Look for high fibre snack bars in the cereal bar aisle, or protein yoghurts in the chilled section.
Best fit supermarket categories:
How to do it
Keep retailer availability, aisle context and product pairings up to date. AI cannot recommend what it cannot confidently place.
Some healthy eating questions move close to medical advice. Brands need to tread carefully.
Health condition or allergy specific diet advice accounted for 7.4% of cleaned questions. Shoppers asked about diabetes, IBS, cholesterol, allergies, dietary needs and other sensitive areas.
Useful partners could include:
Expert reviewed content gives AI safer material to use. It also helps the brand avoid sounding as if it is making claims it should not make.
A safer AI answer might say: Choose products with clearly stated nutrition information and expert reviewed guidance, especially if managing a specific dietary need.
Best fit supermarket categories:
How to do it
Use credible experts where the claim needs support. Be clear where the product can help, and where shoppers should seek professional advice.
Best-fit supermarket categories:
Free-from; Gut health; Low sugar; Plant-based; Sports nutrition.
AI is often better at nutrition logic than appetite.
It regularly returns to:
That is sensible, but shoppers do not buy nutrients in isolation. They buy food they expect to enjoy.
Brands can add what AI often lacks:
Examples:
This is especially relevant to weight management, healthy snacks, swaps and general healthy eating advice.
A flat AI answer says: Choose a lower sugar snack.
A better answer says: Look for lower sugar options that still satisfy the craving for something sweet or crunchy.
Best fit supermarket categories:
How to do it
Do not let health claims crowd out taste. AI needs evidence that the product works in real life, not only on the nutrition panel.
The clearest opportunity sits in the gap between what shoppers ask and what brands currently publish.
AI healthy eating advice is strongest on:
Brands need to give AI more than a claim. They need to give it a role, a use case, a reason to trust the product and a clear connection to the shopper’s real question.
The brands that act first should have an advantage. Not because AI will magically reward them, but because most brand content is still not written for the way people now ask for advice.