A digital public good (DPG) is an open digital resource, such as software, data, an AI system, an open standard, or open content, that helps attain the Sustainable Development Goals (SDGs), complies with privacy and other applicable laws and good practices, and is designed to do no harm. Openness alone does not make something a DPG. The label that carries formal weight is “recognized DPG,” which means a solution has passed review against the Digital Public Goods Alliance (DPGA) Standard and appears in the DPG Registry.
Where the definition comes from
The concept has two layers. The United Nations Office for Digital and Emerging Technologies sets out the broad framing: digital public goods are open-source software, open data, open artificial intelligence models, open standards, and open content. Those resources must adhere to privacy and other applicable laws, standards, and best practices, do no harm, and help attain the SDGs.
As an Amazon Associate I earn from qualifying purchases.
The DPGA turns that framing into something that can be tested. Its FAQ gives a shorter operational definition:
Recommended Free Tools
“Digital public goods (DPGs) are open source software, open data, open AI systems, and open content collections that adhere to privacy and other applicable laws and best practices, do no harm by design, and help attain the Sustainable Development Goals (SDGs).”
The two lists are not worded identically. The UN framing explicitly names open standards, while the DPGA’s short definition lists software, data, AI systems, and content collections. The difference reflects framing versus operational scope rather than a contradiction. Open standards are part of the UN’s policy description, and the DPGA applies its review to the categories it names in its FAQ.
Three terms that are not interchangeable
Writers often use “digital public good” to mean any useful open project. The distinction below keeps the terms precise.
| Term | What it means | What it does not mean |
|---|---|---|
| Open digital resource | A broad description of openly licensed software, data, models, standards, or content. | Any claim of public benefit, SDG alignment, or safety. |
| Digital public good (DPG) | A more constrained public-interest concept: an open resource that meets the SDG, legal, and do-no-harm conditions in the UN and DPGA framing. | An official status that has been granted. Meeting the concept is not the same as passing review. |
| Recognized DPG | A solution that has passed DPGA review against the DPG Standard and is listed in the DPG Registry. | A certification of every local installation, or a permanent status. |
What the DPG Standard tests
The DPG Standard sets the baseline a nominated solution must meet. As accessed in 2026, the Standard page organizes its requirements into nine indicator groups. The indicator areas are:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- SDG relevance: a clear link to one or more Sustainable Development Goals.
- Open licensing: an open license that allows use, adaptation, and sharing.
- Clear ownership: identifiable ownership and governance of the solution.
- Platform independence: the solution is not locked to a single proprietary platform.
- Documentation: documentation sufficient for others to understand and use the solution.
- Non-PII data extraction: the solution does not extract personally identifiable information.
- Privacy and applicable laws, standards, and best practices: compliance with privacy law and recognized practice.
- Data privacy and security: a harm-related safeguard covering how data is protected.
- Inappropriate or illegal content and protection from harassment: harm-related safeguards against harmful content and abuse.
Asset type, license, SDG relevance, ownership, data handling, and harm safeguards are the axes that separate one candidate from another. They describe how solutions are compared under the Standard. They do not show that every solution carries the same risk or delivers the same impact in practice.
How recognition works
Recognition follows a defined sequence, according to the DPGA’s application FAQ:
- Nomination. The solution is submitted for review through the DPGA application process.
- Two reviewers. Each application is reviewed by two individual reviewers, who may refer the solution to external experts.
- Pass or fail. A solution is marked as a DPG if it passes all indicators. It is marked ineligible if it fails one or more.
- Registry listing. Recognized solutions are listed in the DPG Registry, where they can be found.
- One-year validity. DPG status lasts one year and is automatically reviewed at the end of that period.
- Expiry. A solution that no longer meets the Standard may be marked expired and removed from the registry.
Because status is time-bound, a solution listed in one year should be checked against the current registry before anyone describes it as recognized.
Rank #4
Design, not local deployment
The DPGA states that the Standard “exclusively focuses on the design of the core solution,” and that the Alliance “does not evaluate local implementations as part of our review to determine if a digital solution is a DPG.” A recognized DPG therefore tells you that the underlying solution met the design and governance indicators. It does not certify how a particular government, company, or NGO configured, hosted, or operated it. Anyone evaluating a deployment still has to check the implementation’s data handling, legal basis, and safeguards separately.
What the label offers, and what it does not
The DPGA describes several benefits of registry listing: discoverability and promotion, showcasing relevant solutions to stakeholders who may be looking for deployment opportunities, and connecting owners to a global community. These are opportunities the Alliance describes. Listing does not guarantee adoption, funding, measurable impact, or revenue.
The AI context
A UN Office for Digital and Emerging Technologies announcement dated 25 June 2026 described an assessment of AI systems as digital public goods. The announcement says AI systems require analysis of training data, model weights, compute infrastructure, deployment settings, and ongoing governance. It also reported that few AI systems had been formally recognized under the standard in force at that time, and argued that openness is multi-dimensional and does not by itself ensure public benefit or SDG alignment. That count is a snapshot from June 2026. For the full findings, consult the report itself rather than the announcement summary.
Quick Recap
Checks before calling something a DPG
- Confirm the solution appears in the DPG Registry with a current status, not only on the owner’s website.
- Read the recognition date and validity period, since status lasts one year.
- Separate the solution’s design from any specific installation you are evaluating.
- Do not treat “free,” “open source,” or “publicly available” as proof of recognition.
- Check the license and documentation directly, since both are assessed indicators.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
