Search across all content
A proof of concept testing whether AI could help analysts find environmental requirements buried in real procurement documents, tested on EU and Dominican Republic tender data.
Governments increasingly want to buy greener goods and services, but it is difficult to measure whether sustainability policy is actually appearing in tender documents. Environmental requirements are often buried in long technical specifications, written inconsistently, or expressed differently across sectors, countries and languages. Public procurement is estimated to be responsible for around 15% of global greenhouse gas emissions, so better monitoring matters.
The team developed the Green Cure proof of concept to test whether AI could help analysts identify green requirements in real procurement documents. The prototype was informed by Lithuania and tested using European Union Tenders Electronic Daily notices and Dominican Republic procurement specifications. Furniture was chosen as the test category because it is commonly procured and has well-documented European Union green criteria. Lithuania was an important reference point following discussions with the Lithuanian Public Procurement Agency, but the analysis ultimately used European Union Tenders Electronic Daily notices and Dominican Republic procurement specifications because Lithuanian data could not be used as expected.
The prototype showed promise as an information-access tool, especially with a human in the loop. It helped test whether AI could identify green requirements in real procurement documents and highlighted where policy language, procurement documents and available data do not always line up.
AI can help governments monitor green procurement, but it cannot compensate for unclear criteria, poor data or weak oversight. Human review remains important, especially when interpreting technical procurement language across multiple languages and jurisdictions.
The prototype drew on European Union green public procurement criteria, Common Procurement Vocabulary codes, European Union Tenders Electronic Daily notices and procurement specifications from the Dominican Republic.





Connect with 500,000+ public servants solving your hardest challenges.





Connect with 500,000+ public servants solving your hardest challenges.
Help public servants worldwide learn from your work, what worked, what flopped and what you'd do differently
Share your project
Log in or sign up to continue the conversation