In an increasingly globalized and complex food supply chain, traditional supplier evaluation methods are no longer enough. New research, recently published in the journal Foods, marks a major breakthrough by introducing an AI-ready framework designed to predict supplier risk based on real-world global data. We are proud that SGS Digicomply’s trusted datasets played a central role in making this innovation possible.
The Urgent Need for Smarter Supplier Risk Management
According to the World Health Organization, more than 600 million people fall ill each year from contaminated food. With supply chains stretching across multiple countries and regulatory systems, identifying reliable suppliers early has become a global imperative. Traditional methods based on periodic audits and manual reviews are slow, costly, and often outdated before corrective action can be taken.
Recognizing this gap, researchers Sina Röhrs, Sascha Rohn, Yvonne Pfeifer, and Anna Romanova developed a new, dynamic approach—powered by AI and real-time data streams like those found within SGS Digicomply.
Inside the New Risk Assessment Framework
The new supplier evaluation model introduces a structured, quantitative method built around seven critical risk indicators:
Hazard risk
Commodity vulnerability
Incident category
Audit performance
Logistic Performance Index (LPI)
GDP per capita
GDP growth
Each metric is normalized onto a comparable scale, weighted based on its relative importance, and then aggregated into an Overall Risk Score (ORS) for each supplier. This flexible structure allows businesses to customize risk assessments based on their specific operational needs or regulatory environments.
Importantly, the model is designed for AI integration, meaning it can process dynamic updates from live data sources like SGS Digicomply and public alerts such as the Rapid Alert System for Food and Feed (RASFF). If a supplier experiences a product recall or their operating country's logistics performance drops, their risk profile adjusts almost immediately—offering businesses real-time visibility into emerging threats.
What Sets This Approach Apart
This framework overcomes several limitations that have long plagued supplier risk management:
Automation: Risk scoring is no longer manual but automated, enabling consistent, repeatable, and timely evaluations.
Data Flexibility: If audit data is missing—a common real-world issue— the model redistributes weight across available metrics, ensuring a full risk profile is still generated.
Adaptability: Whether operating in high-transparency markets or regions with limited data availability, the framework remains effective.
Continuous Pressure for Improvement: Knowing that scores may be monitored by buyers or made public, suppliers are incentivized to prioritize compliance, traceability, and overall food safety culture.
Validated in the Real World
The researchers manually tested the model on 11 real-world suppliers spanning sectors such as beverages, dairy, and processed foods. Their findings were clear: suppliers linked to major incidents like E. coli outbreaks, fraudulent production, or allergen mislabeling consistently scored in the moderate to high-risk categories. This strong alignment between predicted risk and historical incidents confirms the model’s practical viability.
Unlocking the Future of Predictive Food Safety
Beyond supplier screening, this framework opens the door to true predictive analytics at scale. Embedded within comprehensive platforms like SGS Digicomply—which already manages millions of food safety entries across more than 160 countries—the model could automatically suggest alternate vendors, flag emerging high-risk regions, and simulate procurement decisions based on changing global conditions.
Looking ahead, researchers envision even broader applications, such as integrating environmental data like weather patterns or sustainability metrics, enabling businesses to align food safety practices with broader ESG goals. Climate factors, for instance, could predict risks like mycotoxin growth, enhancing the precision of risk forecasting even further.
The Road to Global Standardization
The study also raises an important point for regulators: voluntary adoption is valuable, but real impact requires a common, harmonized approach. Standardizing AI-driven supplier risk assessments could allow mutual recognition between countries, easing cross-border trade and raising the global standard for food safety.
SGS Digicomply: Empowering the Next Generation of Food Safety Innovation
At SGS Digicomply, we are committed to providing the high-quality, globally harmonized data needed to fuel breakthroughs like this one. Our regulatory intelligence platform ensures that businesses, researchers, and regulators alike have access to the structured, actionable information they need to drive smarter, safer food systems.
This research validates our vision: the future of food safety will be predictive, dynamic, and data-driven. And with the right tools, it’s a future that’s already within reach.
Interested in learning how SGS Digicomply can transform your food safety strategy? Explore our platform here.