GenAI and blockchain in agriculture

Bai, S., Liu, Z., Wu, W., Xu, N., & Jiang, M. (2025). “GenAI + blockchain” to coordinate agricultural supply chains to improve quality trust: An agent-based simulation study. Frontiers in Sustainable Food Systems, 9. https://doi.org/10.3389/fsufs.2025.1591350

Field of application: Agriculture

• Agricultural supply chains specialising in perishable goods management, where leading agribusinesses implement integrated technology platforms to coordinate quality assurance across entire supply chain networks.

Key applications

• Quality detection and monitoring: GenAI technology enables automated quality assessment of agricultural products through intelligent detection systems that analyse appearance, size, colour, and other attributes against established quality standards.

• Freshness prediction and traceability: Blockchain technology provides real-time freshness monitoring and predictive analytics, combined with transparent traceability throughout the supply chain network.

• Inventory optimisation: The integration enables dynamic adjustment of minimum safety stock levels through real-time inventory tracking and demand forecasting, reducing waste whilst maintaining product quality.

• Supply chain coordination: Leading enterprises deploy technology platforms that enable distributors, retailers, and consumers to access quality assurance services, enhancing transparency and consumer trust.

Research methodology

This investigation employs an agent-based simulation framework using AnyLogic software rather than empirical field trials. The computational model compares three distinct scenarios: a baseline model without technological integration, a blockchain-coordinated model focusing on freshness monitoring, and a collaborative model integrating both GenAI and blockchain technologies.

Technology integration framework

• Leading enterprise role: Major agribusinesses serve as technology platform providers, enabling other supply chain members to utilise integrated quality assurance systems rather than implementing individual technological solutions.

• Blockchain capabilities: Provides freshness monitoring through predictive analytics with accuracy parameter α ∈ [0,1], where α = 0 represents no blockchain implementation and α = 1 indicates highly accurate freshness prediction.

• GenAI capabilities: Offers quality detection services with accuracy parameter β ∈ [0,1], enabling automated assessment of product quality attributes and supporting consumer trust development.

• Consumer willingness to pay: The model incorporates consumer sensitivity to freshness and quality attributes, with parameter r representing the strength of quality trust effects on purchasing behaviour.

Implementation considerations

• Technology costs: Implementation expenses scale quadratically with accuracy levels (α + β)², reflecting realistic AI model training requirements where higher precision demands increased data annotation and computational investment.

• Accuracy trade-offs: Beyond precision levels of 0.6, revenue growth rates decelerate, suggesting diminishing returns from excessive accuracy investments. The research demonstrates that pursuing maximum technological precision may not optimise cost-effectiveness.

• Safety stock optimisation: The simulation reveals that minimum safety stock levels of approximately 20-30 units (model-specific parameters) trigger substantial revenue growth whilst balancing inventory costs against market demand fluctuations.

Supply chain performance implications

• Inventory management: The GenAI + blockchain integration achieves more stable inventory levels and precise fluctuation cycles compared to baseline models. Enterprises can maintain reduced safety stock levels whilst preserving product freshness through enhanced demand forecasting.

• Revenue enhancement: The collaborative technology model demonstrates the most pronounced enterprise revenue growth, despite higher initial implementation costs. When GenAI detection accuracy exceeds critical thresholds, enterprises experience exponential revenue growth through freshness premium capture.

• Network coordination: Technology integration enables real-time alignment between consumer demand and production capacity, reducing waiting periods and inventory accumulation across all supply chain participants.

Economic and financial implications

• Cost-benefit analysis: Whilst technology integration requires substantial initial investment, sustained post-implementation revenue growth indicates successful market positioning for freshness-sensitive consumer segments.

• Premium pricing opportunities: Enhanced product freshness assurance, validated through transparent blockchain traceability and GenAI quality detection, enables enterprises to command premium pricing from quality-conscious consumers.

• Operational efficiency: The synergistic application reduces logistics costs, minimises transportation damage risks, and streamlines quality control processes across the entire network.

Research implications for practice

Recent agent-based simulation research demonstrates how generative AI and blockchain integration can fundamentally transform agricultural supply chain coordination for perishable goods. The study reveals that strategic optimisation of minimum safety stock parameters (approximately 20-30 units within the modelled framework) can trigger substantial revenue improvements whilst maintaining product quality and consumer confidence.

Although these findings emerge from simulation-based analysis rather than field implementation, they illuminate critical strategic insights for agricultural enterprises:

• Technology platform strategy: Leading agribusinesses should consider implementing integrated technology platforms that enable network-wide quality assurance rather than pursuing isolated technological solutions.

• Precision calibration: Investment in quality detection and freshness monitoring technologies proves economically viable when calibrated against consumer willingness to pay premiums. However, pursuing maximum technological accuracy may not represent optimal resource allocation.

The research validates findings from real-world implementations, including Charoen Pokphand Group's blockchain and GenAI deployment for durian maturity prediction, which increased accuracy from 50% to 91% whilst reducing safety stock requirements. Such simulation-based research provides the agri-food sector with quantitative frameworks for evaluating technology integration strategies before committing substantial implementation resources, enabling more informed decision-making regarding digital transformation initiatives.