Machine and deep learning for food integrity: A review of emerging analytical applications in safety, quality, and authenticity

The rapidly increasing global population, changing dietary habits and decreasing natural resources are affecting the sustainability and reliability of global food systems. In this context, making food systems resilient to change and sustainable in the long-term requires the adoption of innovative approaches. The development, productization, and quality control processes of food systems, together with data generation, necessitate moving beyond traditional analytical approaches to evaluate food safety and quality. In this regard, artificial intelligence (AI), machine learning (ML), and deep learning (DL) methods are notable for their transformative potential in food chemistry. In recent years, the literature has focused on integrating spectroscopy, imaging systems, electronic nose and tongue sensors, chemical composition and genomic data with AI models, offering solutions to critical problems such as pathogen detection, microbial risk assessment, quality classification, shelf-life estimation, and food fraud detection. These approaches enable the extraction of meaningful patterns from complex food matrices by providing high accuracy, speed, and scalability. The conceptual framework underlying this review reveals that AI-powered analytical food chemistry applications form an integrated structure through predictive systems, automated inspection, quality assessment, and traceability components. Furthermore, food composition and chemical compound databases, along with big data infrastructures, play a critical role in the training and generalizability of AI models. Overall, AI/ML/DL-based approaches offer a new paradigm in food safety management through real-time monitoring, non-destructive analysis, and dynamic decision support mechanisms; however, challenges such as data standardization, model transparency, and regulatory compliance stand out as key issues to be addressed in the future.