Send message to Author

Paper short abstract

Adaptive planning supported by AI can improve food system resilience by enhancing forecasting, land use, and decision-making. Using diverse data and models, this study identifies key success factors and calls for richer data sources to address gaps in uncertainty and public sentiment.

Paper long abstract

Development planning plays a key role in supporting food security and sustainable food systems, particularly in developing countries where governments lead policy efforts. However, such policies often struggle to meet their objectives due to limited data and insufficient consideration of uncertainty. External pressures—such as climate change, economic shocks, and conflict—add further strain, highlighting the rigidity of traditional planning methods. This study explores the potential of adaptive policy frameworks, supported by artificial intelligence, to enhance the resilience of food systems. AI-based tools can support more informed decision-making by improving agricultural forecasting, mitigating soil sealing, and optimising land use and land cover management. An adaptive system using real-time data and predictive modelling is proposed to help policymakers respond more effectively to emerging risks. A combination of analytical methods—text mining, regression analysis, Random Forest, and artificial neural networks—is used to develop the system. Data sources include the International Aid Transparency Initiative, Global Food Security Index, World Uncertainty Index, world news APIs, and National Development Plans. Results suggest that AI-supported approaches can help identify key project factors—such as budgets, timelines, and national priorities—associated with more successful outcomes. The AI-driven adaptive system also shows strong potential to support evidence-based development planning and decision-making. Nonetheless, challenges remain. Data gaps and the limited availability of structured information on public sentiment and uncertainty can affect model performance. Future work should explore alternative data sources—such as social media and additional development indices—test other algorithms to improve predictive accuracy, and conduct thematic deep dives to generate more targeted policy insights.