Researchers Test if AI Can Create Burgers Better Than a Big Mac

On November 16, 2026, researchers at Stanford University and the Stanford University School of Medicine revealed a generative artificial intelligence model trained on 2,216 human-written burger recipes that successfully created novel patties tested against one of the world’s most recognizable fast-food items, News-Medical reported.

Traditional approaches to food creation rely heavily on culinary skill and iterative trials, which can slow the search for new formulations that balance taste, nutrition, and environmental impact. Food production contributes to climate change, biodiversity loss, and freshwater depletion, while inadequate dietary patterns result in malnutrition and chronic diseases. Sustainable alternative foods face a primary constraint in consumer acceptance, which depends heavily on taste, texture, and cultural familiarity.

The research team implemented a two-stage generative AI model trained on a curated collection of over half a million recipes on Food.com. The dataset covered 146 ingredients, teaching the model which ingredients commonly appeared together and in what quantities. The AI generated one million burger recipes and assessed them for novelty, predicted palatability, environmental impact, and nutritional quality.

Training the model on millions of recipes

While ingredient-quantity estimates were less precise, the generated samples closely matched the original dataset regarding ingredient popularity, ingredient correlations, and recipe length. Popular ingredient combinations generally occupied regions associated with higher predicted palatability but lower nutritional scores. In a separate test, the model produced a recipe that aligned with a reconstructed Big Mac reference under the study’s similarity criteria, even though it had not been trained on that specific reference.

The researchers assembled the reference from four publicly available recreations because the official formula is proprietary. Across ten independent runs, finding a match required an average of 7.3 million generated recipes, allowing some variation in ingredient quantities without establishing that the model reproduced the exact commercial formula.

Evaluating taste and environmental impact

An executive chef translated the ingredient lists into standardized cooking and assembly instructions for a blinded sensory evaluation involving 101 volunteers at a restaurant in San Francisco, California. Every volunteer sampled all six burgers, providing ratings for overall liking, flavor, and texture on a scale from one to seven.

Delicious Burger 1 had slightly higher flavor ratings than the Big Mac, while Delicious Burger 2 had slightly higher overall liking and flavor ratings, with texture scores remaining similar. Participants more often described Delicious Burger 1 as meaty, moist, and fatty, whereas Delicious Burger 2 was described as meaty and smoky.

For environmental performance, the team selected two mushroom-containing burgers. Achieving an environmental impact score of 0.06, Sustainable Burger 1 (a mushroom-based formulation) sat roughly 94% lower than the Big Mac score of 0.93. Sustainable Burger 2, combining beef and mushrooms, registered an environmental impact score of 1.02, which came in slightly higher than that of the Big Mac.

Sensory evaluation showed that the mushroom-based patty received significantly lower ratings for liking, flavor, and texture compared to the conventional burger, while Sustainable Burger 2 received similar ratings with no statistically significant differences detected.

Nutritious burger improves health scores but tastes worse

The model also produced a Nutritious Burger with a Healthy Eating Index (HEI) score of 63.12, nearly twice the Big Mac’s score of 33.71. Its estimated environmental impact score was about one-sixth of the Big Mac’s.

Derived from ingredient data rather than proven via clinical health outcomes, the formulation incorporated increased amounts of vegetables, whole grains, and plant protein alongside decreased levels of refined grains, sodium, and saturated fat. During tasting, the nutritious burger scored significantly lower on overall liking, flavor, and texture than the conventional burger, with participants describing it as earthy, bland, dry, soft, and grainy.

The model also produced personalized burgers based on age, sex, body composition, and physical activity, such as recipes for an active 15-year-old male and a moderately active 70-year-old female. These examples were computational and were not included in the restaurant tasting.

An evaluation across all six burgers revealed clear trade-offs balancing novelty, ingredient diversity, sensory acceptance, nutrition, and ecological footprint. The results show that generative AI can use publicly available recipes to create burgers selected for different goals, though no single tested burger achieved the highest scores for taste, nutrition, and sustainability at once.

Photo of author

Sophie Lin - Technology Editor

Sophie is a tech innovator and acclaimed tech writer recognized by the Online News Association. She translates the fast-paced world of technology, AI, and digital trends into compelling stories for readers of all backgrounds.

West End Star Casting Reliance Risks Stunting Audience Intellect