A laboratory result can identify an unwanted substance in food without answering the question that matters most to a producer: what should happen to the batch? That decision depends on the substance, its concentration, the reliability of the measurement, the intended consumer and the applicable rules. Faster detection is useful only when those connections are understood. Europe’s food safety challenge is therefore partly an organisational one. Laboratories, manufacturers and authorities need information that can move from a test report into a practical decision. Better sensors and analytical methods can support this work, but their performance must be judged alongside sampling, traceability and the ability to prevent a problem recurring.
Distinguishing detection compliance and risk EFSA’s latest pesticide-residue assessment illustrates why terminology matters. In the EU-coordinated programme, 9,842 samples collected in 2024 covered a specified selection of commonly consumed foods. Some 98.8 percent complied with EU legislation. Residues exceeded maximum levels in 2.4 percent while 1.2 percent were confirmed as noncompliant after measurement uncertainty was considered. These figures describe that sampling programme, not every food sold in Europe. A detected residue is not automatically an unlawful residue, and a regulatory exceedance is not itself a measurement of illness. EFSA separately assesses dietary exposure and health risk.
Equally, a compliant result for the substances tested should not be presented as proof that a product contains no other hazards. The question asked of the laboratory determines what its answer can establish. For manufacturers, this distinction should shape the information supplied with each result. The report needs to identify the sample, method, reporting limits and uncertainty where relevant. A purchasing team receiving only a green compliance symbol may lose the information required to compare suppliers or investigate an unexpected change. A trial of a new method should therefore specify the decision it is intended to support before comparing speed or price. Releasing a finished batch, investigating a supplier and monitoring a process can require different levels of certainty.
Combining those purposes in one performance claim makes it harder for buyers to judge whether the technology answers their particular question. Sampling needs to match the production decision A test measures the material submitted, rather than the entire production history behind it. This creates a practical research question: how should a sampling plan reflect changes in raw materials, suppliers, storage and processing? Testing more frequently is not necessarily the same as learning more about the most important sources of risk. Consider a hypothetical cereal processor introducing a new grain supplier. A sample taken from one delivery cannot describe every subsequent delivery or the conditions in every storage silo.
The useful comparison would connect analytical findings with the origin, storage history and processing conditions of the material. Those details help determine whether an unusual result is isolated or part of a pattern. EU hygiene rules provide the underlying framework for managing food safety throughout production and distribution. New testing technologies need to support those controls, rather than become a substitute for them. A successful demonstration should explain how the result changes an operating decision, who receives it and what happens while uncertainty is resolved. Traceability must remain usable when something goes wrong.
The General Food Law places primary responsibility for food-law compliance with businesses. Its traceability framework generally requires operators to identify their immediate supplier and immediate business customer, with specific rules for some sectors. Traceability supports targeted withdrawals and recalls when required; it does not make an unsafe product safe. The operational test is whether information can be retrieved and interpreted quickly. Ingredient records, production batches and dispatch information may each exist but use different identifiers. Connecting them matters especially where one ingredient enters several products or a finished batch reaches multiple distributors.
This gives digital innovation a concrete purpose. A traceability project could measure the time required to reconstruct a batch’s route, the proportion of records that can be matched and the precision of a simulated withdrawal. Such measures are more informative than counting how many records a platform stores. They reveal whether the system can help people act under pressure. A mock recall should also test responsibility. Staff need to know who can stop dispatch, escalate an uncertain result and contact customers or the relevant authority. Recording where the exercise stalls can reveal missing handovers or unclear decision rights.
The useful outcome is an agreed correction that is tested again, rather than a completed drill that leaves the same organisational weakness in place. Food fraud requires more than an unusual test result. Authentication technologies can help identify unexpected ingredients or inconsistencies with a product description. However, an analytical anomaly does not establish deliberate fraud. The Commission distinguishes suspicions of agri-food fraud through criteria involving a rule violation, intentional conduct, deception.
