Integrated Artificial Intelligence Blockchain Internet of Things Model for Improving Transparency in Seed Supply Chain in Kenya

Seed sector plays a vital role in advancing agricultural sector. Lack of transparency has been highlighted by previous studies as the leading challenge in seed supply chain that has made farmers have limited access to reliable and sufficient seed information. This has led to purchase of counterfeit seeds resulting in crop failures, financial losses and low productivity among others. This study aimed to analyze data and develop an integrated AIBIoT model to enhance transparency seed supply chain. The objectives of the study were to assess the digital transparency indicators, evaluate AIBIoT capabilities in seed supply chain, evaluate the factors that affect the implementation of AIBIoT, and to develop an integrated AIBIoT model for improving transparency in seed supply chain. The study adopted pragmatism research philosophy with mixed methods descriptive-causal survey research design having both qualitative and quantitative data. Data was collected from KALRO, KSC, agro-dealers, and farmers through questionnaires and interviews. Quantitative data were analyzed using descriptive and inferential statistics, including correlation and regression analyses, supplemented by factor analysis while qualitative data were analyzed using thematic analysis. Findings showed that there were very low indicators of digital transparency though there was a positive perception among the stakeholders on AIBIoT capabilities with significant variability in usage confidence and implementation. In addition, statistical tests showed that AIBIoT technologies have a positive significant statistical effect on transparency with likelihood ratio Chi-Square test of (χ² = 28.168, df = 13, p = 0.018). Implementation challenges were significantly affecting the effectiveness of AIBIoT with regression results (R = 0.611), with 36.4% of the variance in adoption explained by challenges (R² = 0.364). Significant predictors of lower of AIBIoT were high cost of implementation (p = 0.034), limited technical knowledge (p = 0.021), lack of awareness of technology benefits (p = 0.030), reluctance to transition (p = 0.048), and concerns about cost and learning curve (p = 0.020). Developed model was found to be statistically significant with p = 0.030, meaning it significantly predicts the dependent variable better than the intercept-only model because it indicates that when AIBIoT moves from 0 to 1 that is, when it increases by 1 unit, the probability of higher transparency increases from 40.61% to 42.23%, an increase of 1.62%.  It was, however, moderated by implementation challenges with statistical results showing negative coefficients associated with interaction terms. Recommendations include enhancing digital transparency, focusing adoption on highly performance technologies, enhancing interventions on high implementation challenges and adoption of IAIBIoT model to improve transparency in seed supply chain. Contributions were to the body of knowledge through contextual understanding of transparency gaps, enriching scholarly studies on AIBIoT implementation challenges and developing a novel theoretical model that can be adapted, validated further or compared. It contributes to practice by providing a transparency diagnostic road map, contextual technology capability guidance and strategic planning guidance on necessary investments and interventions needed to curb identified barriers of AIBIoT implementation.