Innovating traceability with stable isotopes, multi-elemental analysis, and explainable AI: validated using saffron as a commodity

Food fraud poses a significant global challenge, with an estimated impact on around 1% of the international food industry, resulting in annual losses of approximately $40 billion[1]. The complexity of global supply chains, which often prioritize speed over quality, exacerbates the issue and drives the demand for greater transparency and authenticity in the food supply chain. To address this issue, a range of analytical techniques—including elemental composition analysis, isotopic methods, spectroscopy, chromatographic and molecular techniques—are employed to verify the authenticity of foods. In addition to these physical testing methods, digital tools such as artificial intelligence (AI) and machine learning are revolutionizing food traceability. By analysing complex datasets, AI algorithms deliver predictive insights that help identify potential fraud and significantly enhance the accuracy of authenticity verification.