Discover organizations working across AI and learning, explore how they are connected, and find reusable public goods for your next investment, program, or policy decision, especially for the Low & Middle Income Country (LMIC) contexts.
Frameworks, benchmarks, datasets, content, standards and decision tools for AI in learning, all free to use today.
A public good is a free, ready-to-use resource that helps others develop, improve, evaluate or implement AI in education. It can be technical (like a dataset, AI model or software tool) or non-technical (like a framework or implementation guide). What matters is that other organisations can use it or build on it for their own work, beyond the project it was originally created for.
Your resource should meet all four criteria:
| Likely to qualify | Unlikely to qualify |
|---|---|
| Datasets, AI models and benchmarks available for others to use in developing or evaluating solutions | Datasets or model demonstrations that cannot be reused in others’ work |
| Software components, free APIs and technical infrastructure that others can integrate or build upon | Finished AI applications without reusable components |
| AI procurement frameworks, evaluation methodologies and implementation guides with practical steps others can apply | High-level project descriptions without transferable methods or guidance |
| Curriculum standards, competency frameworks and learning taxonomies that can inform other AI-enabled solutions | Individual lesson plans or teaching materials intended only for classroom use |
| Reusable frameworks, tools or methodologies developed through research | Research papers or articles that present findings without a reusable resource |
Good to know: Resources do not need formal Digital Public Goods certification or fully open-source licensing to be included. The Library is a discovery platform, not a certification registry. Inclusion does not imply independent verification of a resource’s quality, safety, effectiveness or legal compliance.
The file is view-only. To add or update an entry, or to share feedback, use Contribute: changes made to a downloaded copy do not update the website.
The file includes every column in the dataset. Some are left off the website for brevity, such as identifiers, headquarters coordinates and the sources behind key initiatives. They are kept in the file to support analysis.