DPGs for AI Collection
Advancing trustworthy, accessible, and equitable AI solutions for the public interest
Digital public goods offer a transformative pathway to democratise artificial intelligence and reduce the global AI divide. By promoting open-source software, open AI training, fine-tuning and benchmarking datasets, open content collections, and open AI models, DPGs lower barriers to AI adoption, reduce systemic biases, advance digital autonomy and broaden opportunities for all.
 Open Data for Public Interest AI.png)
The DPG4AI collection aims to:
Elevate Visibility
Amplify DPGs driving the ethical and responsible development of AI to support the SDGs in an equitable way.
Illustrate Impact
Feature adoption stories and best practices showing how DPGs enable responsible AI development.
Map the Ecosystem
Provide an overview of AI-relevant DPGs to help navigate the open-source ecosystem, especially in low-resource contexts and underrepresented domains.
Co-stewarded by the United Nations International Computing Centre (UNICC), the Open Data Initiative (ODI), Digital Futures Lab and the DPGA Secretariat, the DPG4AI Collection is a curated set of digital public goods that can be utilised at different stages of the AI development lifecycle, are grounded in real-world needs, and have a documented impact. Particular attention is given to the tools' relevance for low-resource contexts and underrepresented domains, allowing local developers and governments to develop and adapt AI solutions to their specific cultural, linguistic, and socioeconomic contexts. By doing so, we aim to ensure that AI is developed safely and securely, with communities at its core, and to create opportunities for everyone across domains such as healthcare, education, and agriculture.
Building on the global momentum for responsible AI governance, this collection contributes to translating the commitments of the UN Global Digital Compact, the UN Secretary-General's High-Level Advisory Body on AI, and the UN Independent Scientific Panel on AIinto practice. It also supports and complements the efforts of members and partners, such as the UNDP AI Hub for Sustainable Development.
The collection serves as a vital repository for governments, civil society, and multilateral institutions seeking to discover and utilise trusted, open source tools to drive responsible and equitable AI development for the public interest. It also provides a starting point for funders, researchers, and developers to identify high-impact tools that should be sustained and scaled globally.
How DPGs are identified: Collection criteria for responsible and equitable AI
To ensure the solutions in this collection are safe, secure, relevant for low-resource contexts and adaptable by developers in resource-constrained environments, the DPGA Secretariat—in collaboration with UNICC, ODI and Digital Futures Lab, as well as based on community feedback—developed a set of criteria built on the DPG Standard for identifying suitable solutions for this collection.
DPG4AI Collection CriteriaDigital Public Goods
Join the DPG for AI Collection
The DPG for AI collection is updated on a rolling basis as new software solutions, datasets, models and applications are submitted and vetted. As the open source AI ecosystem grows— particularly in edge computing, federated learning, and localised training and testing datasets—the collection will expand to reflect these innovations.
Do you represent a current DPG that you feel should be included? Email us for more information.
Email usIf you work on a solution that is not a DPG but would like to be included within the collection, start by applying to become a DPG.
Apply to become a DPGResources
