Harms screening for tech in development

Tech Harms Check is a screening tool and digital framework that evaluates the human rights implications of technology use in development and humanitarian projects. It is designed for non-expert practitioners, including program officers and civil society staff, who can upload a project proposal and receive a structured assessment identifying potential harms, severity scores, and tailored mitigation strategies. AI is used to simplify the user's interaction with the framework by extracting relevant technological features and contextual factors from the proposal and connecting them to a curated library covering 54 human rights across multiple scoring dimensions.

A core design principle is the strict separation between database-driven content, which flows through verbatim from the curated indicator library, and the limited use of generative AI, which is confined to narrative summaries and contextual mitigation suggestions and is clearly marked wherever it appears.

Draftlab and Dot Studio partnered on the implementation of the MCP server that connects the curated library to the assessment workflow. The work spans translation between the framework and a working data model, UX for non-experts, and a business strategy for keeping the tool sustainable past launch.