YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

CABI has published its first artificial intelligence‑ready datasets on Hugging Face, a global platform for sharing machine learning models, datasets, and AI resources. The datasets were developed under the Generative AI for Agriculture (GAIA) project, which aims to make agricultural knowledge more accessible to developers building AI tools.

The GAIA project is supported by the Gates Foundation and UK International Development from the UK government. Its objective is to address inequalities by improving content sourcing, sharing, and governance across generative AI systems, with the goal of making digital advisory tools safer, more relevant, and more equitable.

Katherine Cameron, Global Lead for Digital Advisory Services at CABI, explained that making agricultural content available through Hugging Face would help developers and researchers integrate trusted, science‑based information into AI applications. She noted that demand is increasing for AI tools that provide practical, locally relevant agricultural advice, and that AI‑ready formats can improve the accuracy of digital advisory tools and support better decision‑making in the agricultural sector.

The first dataset, CABI Plant Health Knowledge for AI, compiles locally co‑developed resources on crop pests and diseases in Kenya, Ethiopia, and India. It includes PlantwisePlus factsheets, pest management decision guides, and plant health case studies, offering developers locally relevant information to support smallholder farming systems.

Another dataset, CABI Palm Weevil Knowledge for AI, focuses on palm weevils, including the red palm weevil, a major pest affecting palm production worldwide. The pest is estimated to cause approximately US$2 billion in crop losses annually and impacts the livelihoods of about 50 million farmers. The dataset integrates books, journal articles, datasheets, factsheets, and country‑specific information, along with registered bioprotection products where available. It was developed in response to growing interest in AI applications for red palm weevil management, a priority area for the Gates Foundation‑funded AgriLLM initiative.

CABI plans to release additional datasets in the coming months, including a livestock dataset covering poultry and cattle later this year. The Plant Health Content collection will also expand with further resources, including bioprotection product information. Developers and researchers interested in AI applications for agriculture are encouraged to explore the datasets.