The debate over “ultra-processed foods” (UPFs) continues to search for a universal definition. But classification alone is not enough: categories describe a food, not how it interacts with the person eating it. Precision nutrition and artificial intelligence can combine data on food composition, health, eating habits and physical activity, turning it into personalised guidance on portions, frequency and food combinations. Not a one-size-fits-all verdict, but guidance that evolves with each person.
Key Messages
- A category does not know the person behind the plate. Effective prevention considers portions, frequency, health, lifestyle and the overall diet.
- Processing does not mean making food worse. Food processing is a tool: what matters is how it is used and the outcomes it produces.
- Not one label for everyone, but individual guidance for each person. Precision nutrition and AI can turn data into personalised, dynamic and practical recommendations.
What’s Happening
Everyone is searching for a clear definition of UPFs. Yet no one seems able to agree on one. The further the debate advances, the less clear the category’s boundaries become.
Even the U.S. Food and Drug Administration (FDA) now acknowledges the challenge: defining UPFs is proving far more complex than expected because of the many competing classification systems and the still-evolving scientific evidence.
A shared classification can support research, track consumption and make studies easier to compare. But it cannot determine whether a food is right for a particular person. How much and how often it is consumed also matter, as do the person’s overall diet, lifestyle, age, metabolic health and individual goals.
When Classification Becomes Policy
Once a classification enters public policy, its role changes: it is no longer just a tool for analysis, but a basis for government action.
It can become the basis for labels, warnings, taxes, restrictions or mandatory reformulation. A scientific boundary that remains contested can therefore harden into a rigid regulatory divide between foods deemed “good” and “bad”.
Food Processing Is a Tool, Not the Enemy
Food processing is not inherently harmful. It has always been part of how we produce and consume food, helping to preserve it, improve safety, increase accessibility and convenience, enhance key qualities, and drive innovation.
The key question is not whether a food is processed, but how it is processed and with what outcomes. Processing can improve nutritional quality, support portion control and help keep costs affordable.
Food safety and nutritional value are two different issues. A product can meet every safety standard without being suitable for frequent consumption.
From Data to Better Choices: AI-Powered Precision Nutrition
Citizens do not need another label that turns a classification into a list of permitted or prohibited foods, but guidance that can take the context of consumption into account.
The connection is simple: precision nutrition starts with individual needs; artificial intelligence turns this approach into practical, personalised guidance.
A nutritional assistant could connect food composition data with voluntarily shared information on health, habits, lifestyle and goals, providing adaptable guidance on portions, frequency, food combinations and alternatives.
For this model to be reliable and accessible, public policies must establish standards for food data, protect privacy, ensure transparency in the criteria used, and broaden access to technology. AI should complement, not replace, nutrition education, physical activity and the creation of health-promoting environments.
More useful information strengthens autonomy, guides demand and drives innovation. The future of prevention will not impose the same choice on everyone: it will help each person find the one that works best for them.