Bias-Aware AI in Agricultural Extension: Enhancing Equity and Inclusivity Through Human-Centered Design
2025
Nganga, Kevin Gitau | Ghosh, Aniruddha
This Info Note explores the transformative potential of Artificial Intelligence (AI), particularly Large Language Models (LLMs), in agricultural extension services. It emphasizes the importance of inclusivity and bias mitigation to ensure equitable outcomes for smallholder farmers, women, and marginalized groups. While LLMs can enhance climate resilience and decision-making by offering timely, context-aware advisories, they risk reinforcing systemic biases if not carefully designed. Moreover, it advocates for the integration of Human-Centered Design (HCD) principles and participatory methods throughout AI development to align technologies with diverse user needs. A novel methodology using the DALL·E image generation model demonstrates how prompt engineering can mitigate stereotypical representations in AI outputs. By combining ethical AI practices, localized insights, and inclusive visual and textual content, the InfoNote presents a roadmap for equitable innovation in agronomic and climate information systems. Policy and governance recommendations to foster trust, transparency, and broad adoption of AI tools in agriculture are also outlined.
显示更多 [+] 显示较少 [-]