Health & Environment
Study Finds AI Calorie-Tracking Apps Often Inaccurate for Complex Meals
Škoda We Love Cycling1h
THE BRIEF
1
A study by Fridolfsson et al. found that AI-powered food analysis apps exhibit error rates between 48% and 66% when estimating macronutrients in multi-ingredient meals.
2
While apps like SnapCalorie and Cal AI claim high accuracy for simple items, researchers noted a systematic underestimation of portion sizes and difficulty identifying hidden ingredients like oils and sauces.
3
The 2026 systematic review in PMC warns that these inaccuracies make current AI tools unreliable for competitive athletes who require precise nutritional data for fueling strategies.