On 3 November 2026, the AI Hub for Sustainable Development opened applications for AI Diffusion Pathways for Agriculture (AIDPA), a six-month programme that will give one agritech venture in each of seven countries access to a shared AI model built with the Moroccan company DeepLeaf. The programme was launched at the Rome event “AI Diffusion Pathways: The Frontier is Collaboration” at Luiss Guido Carli University. Applications close on 15 December 2026 at 23:59 CET.

The AI Hub is co-led by Italy’s Ministry of Enterprises and Made in Italy (MIMIT) and the United Nations Development Programme (UNDP). DeepLeaf is the programme’s delivery and platform partner. The official programme page lists seven pilot countries: Ghana, Italy, Kenya, Morocco, Rwanda, Tanzania and Tunisia. Applicants must be based in, operate in or show a credible commitment to working in one of them.

The shared model is the core of the offer. UNDP says DeepLeaf’s African-developed model has already been validated in Kenya, Morocco and Italy. StartupNews Africa has not reviewed a validation study; the claim comes from UNDP and has not been independently verified. UNDP also says that solutions built on the model stay locally owned, that agricultural data is shared only with consent, and that open components return to the public as digital public goods.

The sources differ on how many local teams gain access. The UNDP release says local agritech start-ups in each of the seven countries will build their own solutions on the shared model. The programme page describes a narrower design: one venture selected in each country, in a cohort running from February to July 2027. This article follows the programme page, which sets out the application terms.

Selected ventures receive platform access, software development kits and application programming interfaces at no cost until 31 July 2027. They also get cohort and venture workshops, technical assessments, regular progress check-ups, access to pilot sites and introductions to potential customers. Demo Day is scheduled for 3 August 2027.

Platform access is not capital. No funding amount for the programme has been disclosed in the sources reviewed, and no investment in DeepLeaf is disclosed in them either.

The call is open to agritech ventures at any stage, including early-stage ideas. Startups, social enterprises and other organisations may apply, and individuals may be eligible with a compelling plan. Selection criteria include the applicant’s ability to carry out the plan, relevance to local development priorities, access to local agricultural knowledge, product readiness, responsible data use and commercial sustainability. In a 9 November LinkedIn post, DeepLeaf said applicants need no finished product, users or revenue to apply.

DeepLeaf’s StartupList Africa profile describes an AI crop-health company whose flagship model, DeepLeaf-Agronomist-1-4B, covers 83 crops, and lists no disclosed funding rounds.

Open questions include which ventures will be selected, whether the validation evidence will be published, and the model’s licensing terms. The programme page, the UNDP release and DeepLeaf’s post do not set out licensing terms for the shared model.

Sources

TopicsAgritech Artificial Intelligence Morocco Africa