DevelopMetrics' work is grounded in a simple principle: AI should help people make better use of evidence, not replace the judgment required for policy decisions.
That principle is reflected in the peer-reviewed article Integrating human-centered AI for land use policy: Insights from agricultural interventions in international development, published in _Land Use Policy_.
The article, by Lindsey Moore and co-authors, examines how human-centered AI can support agricultural and land-use policy work in international development. It focuses on the practical challenge of helping decision-makers work with complex evidence, including policy documents, intervention records, contextual information, and implementation constraints.
Why this matters
Land-use and agricultural policy decisions rarely depend on one clean source of information. Teams need to consider research, program records, local context, political and institutional constraints, and evidence about what has worked in similar settings. That evidence is often scattered across documents and systems.
AI can help organize and interpret that information, but only when the workflow is designed around human judgment. Decision-makers still need to see where claims come from, what assumptions are being made, and where the evidence is uncertain or incomplete.
How it connects to DevelopMetrics
This is the same operating logic behind DevelopMetrics' evidence systems and AidInsight work. We build tools and workflows that help teams:
- bring fragmented evidence into one usable working frame,
- connect outputs back to source materials,
- make assumptions and limitations easier to inspect,
- and support policy, reporting, funding, and delivery decisions with reviewable evidence.
The article helps show that our approach is not just a product idea. It is part of a broader research agenda on how AI can be used responsibly in development-sector decision-making.
Read the original peer-reviewed article in _Land Use Policy_: Integrating human-centered AI for land use policy.