What information does the AI need about you?

Goals (weight loss, weight gain, stable weight or support for training), body data such as age, gender, height and body weight, activity level, training type and timing, food preferences, allergies and intolerances, diet type such as vegetarian or vegan, and how many meals you want and when.

How are allergies handled differently from food preferences?

Registered allergies and intolerances are linked to structured data about ingredients and allergens, and each recipe's ingredients are checked against what the user cannot tolerate. This matters especially for composite foods such as sauces, broths and spice blends, where the allergen is not always clear from the ingredient name. Recipes containing a registered allergen are excluded, and substitute ingredients are also checked.

Does linking diet to training mean eating more calories?

Not necessarily. The same amount of energy and nutrients can be distributed differently throughout the day, for example by allocating a larger share of the day's carbohydrates before and after long or demanding sessions.

Can the plan be changed without starting over?

Yes. If the user wants a different meal, the system can find an alternative with approximately the same energy and nutrients while still respecting diet type, food preferences, allergies and intolerances. The plan can also be updated if goals, body weight, training volume or eating habits change.

When should an AI diet plan not be used alone?

In the case of illness or other conditions that require medical nutritional therapy, the diet should be monitored by qualified health personnel.

AI Diet Plan: How Your Diet Can Be Customized | Hardworkout Magazine

Key takeaways

- A specialised AI diet system can combine calculations, structured food data and national dietary guidelines, rather than relying on nutritional values the AI generates itself.

- In a 2025 systematic review of 11 studies, six out of nine studies with comparison groups showed statistically significant improvements from AI-based dietary interventions.

- Registered allergies and intolerances can be checked against structured ingredient and allergen data, so recipes containing an allergen are excluded and any substitute ingredient is checked before use.

- Linking the diet plan to training changes when energy is eaten rather than how much: the same total energy and nutrients can simply be distributed differently across the day.

Artificial intelligence can be used for more than just suggesting individual meals or recipes. With an AI meal plan, energy, nutrients, and meals can be tailored to goals, activity levels, training, and food preferences.

In specialized systems, personal information can be combined with food data and nutritional guidelines. The plan can also take into account allergies, intolerances, and foods you want or do not want to eat. This provides far more opportunities for individual customization than when a regular chatbot generates a daily menu.

What is an AI Diet Plan?

An AI diet plan uses artificial intelligence to create a diet based on information about the person following the plan.

The simplest variant is to ask a general chatbot for a diet plan. If you provide the desired calorie intake and number of meals, it can, for example, create a daily menu with foods and recipes. If you add goals, training, and food preferences, it can attempt to further customize the suggestion.

A specialized AI system for diet can be built in a completely different way.

Here, AI can be combined with calculations, structured food data, and nutritional science rules. The system can use Matvaretabellen for nutritional values and build recommendations based on research and national dietary guidelines. Thus, calculations and nutritional content do not need to be based on values that AI generates itself.

A [study from 2025](https://pubmed.ncbi.nlm.nih.gov/40565444/) reviewed 11 previous studies where AI used, among other things, biological measurements and information from users to provide personalized dietary recommendations.

In six out of nine studies with comparison groups, the AI-based interventions showed statistically significant improvements.

How do specialised AI agents divide the work on a diet plan?

In more advanced systems, different AI agents can have their own tasks when the diet plan is being assembled.

One agent can work with energy and nutrients, while another selects recipes that fit the goals and the user's preferences. Others can check for allergies and intolerances or verify the plan against nutritional guidelines.

The suggestions can thus be evaluated in several steps. A recipe may, for example, meet the goals for calories, proteins, carbohydrates, and fats, but still be excluded because it contains an ingredient the user cannot tolerate.

What Does AI Need to Know?

How well the diet can be customized depends, among other things, on what information the system has about the user.

This can include:

- Goals: Weight loss, weight gain, stable body weight, or support for training and performance.

- Body Data: Age, gender, height, and body weight.

- Activity Level: How active the person is throughout the day.

- Training: Type of training, amount, frequency, and timing.

- Food Preferences: Foods and dishes the person likes or wishes to avoid.

- Allergies and Intolerances: Ingredients that must be excluded or replaced.

- Diet Type: For example, vegetarian or vegan.

- Meals: How many meals the person wants and when it is convenient to eat.

How is the energy requirement estimated, and how accurate is it?

The energy requirement is a natural starting point when assembling the dietary plan.

Age, gender, height, and body weight can be used to estimate the resting metabolic rate. Then, activity level and training are taken into account when calculating the total energy requirement. The goal determines whether the energy intake should be approximately at the estimated requirement or if a calorie deficit or surplus should be implemented.

The calculation will still be an estimate. A [study from 2023](https://pubmed.ncbi.nlm.nih.gov/37632665/) reviewed 29 studies with a total of 1430 athletes and compared various formulas for the resting metabolic rate. Even the most accurate formula was within ±10 percent of the measured value for about 80 percent of the participants.

Changes in body weight over time can therefore be used to assess whether the energy intake should be adjusted.

How are protein, carbohydrates and fat distributed?

Once the energy intake is calculated, goals can be set for proteins, carbohydrates, and fats.

The distribution can be adjusted based on body weight, goals, and the type of training the person is undertaking. Someone who trains a lot of strength, for example, may have different needs than someone who performs several long endurance sessions each week.

Nutritional guidelines can be used to set the framework for how much protein, carbohydrates, and fat the plan should contain.

How to Build an AI Nutrition Plan

Once the goals for energy and nutrients are set, they should be distributed across actual meals.

The system can use nutritional data from foods and recipes to select meals and calculate appropriate portions. If the user wants three larger meals, the daily intake is distributed differently than if there are to be five smaller meals.

The recipes and portions can then be adjusted so that the sum of the meals is close to the goals for energy, proteins, carbohydrates, and fats.

![](https://hardworkout.no/wp-content/uploads/2026/10/ChatGPT-bilde-1.-okt.-2026-095129.png)

Food Preferences and Recipes

Two people with the same energy and nutrient needs do not need to have the same meals.

Food preferences can be used to prioritize recipes with foods the user likes and exclude foods the person does not want. This could include a desire for specific types of breakfast, less red meat, or dishes that fit certain times of the day.

Diet type, such as vegetarian or vegan, can also guide which recipes are relevant. If an ingredient is swapped out, the amounts and nutritional content can be recalculated so that the recipe still fits into the plan.

[](https://hardworkout.no/ai-kostholdsplan-slik-kan-kostholdet-tilpasses-deg/)

Nutrition

AI Nutrition Plan: How Your Diet Can Be Customized

Artificial intelligence can be used for more than just suggesting individual meals…

Date: Oct 01, 2026

[](https://hardworkout.no/test-therabody-theraface-mask-ledmaske/)

Electronics

Test: Therabody TheraFace Mask LED Mask

LED masks have become an increasingly popular option for skincare at home.…

Date: Oct 01, 2026

[](/magazine/15-minute-workout-program-for-a-stronger-core)

Free Workout Program

15-Minute Workout Program for a Stronger Core