Open Data Manual › Practical guides › The CC0 License Explained for Data Publishers

The CC0 License Explained for Data Publishers

By Open Data Manual editors · Published 4 November 2026 · Editorial standard

CC0 is the tool most often chosen when a publisher wants to put data as close to the public domain as the law allows. Instead of granting permission subject to conditions, it gives up the rights that would otherwise restrict reuse. This guide explains how CC0 works, why it suits many datasets, and the limits that publishers sometimes overlook. For the wider question of whether data is protected at all, see the manual's appendix on legal rights in data and databases and our guide to database rights.

What CC0 does

CC0 1.0 Universal is published by Creative Commons as a public domain dedication. The person applying it (the "affirmer") waives their copyright and related rights in the work worldwide, to the extent the law permits. Once that is done, anyone may copy, modify, distribute and use the data, including commercially, without asking permission and without conditions.

The legal code works in three layers:

1. Waiver (section 2). The affirmer gives up their rights permanently and unconditionally, for every territory and the longest term the law allows. The waiver is stated to be irrevocable.
2. Public licence fallback (section 3). Some legal systems do not allow authors to abandon all their rights. Where the waiver fails, CC0 grants everyone a royalty-free, irrevocable licence with the same broad scope, and the affirmer promises not to assert any rights that remain.
3. Limitations (section 4). Trademark and patent rights are not waived, the work comes without warranties, and the affirmer does not clear rights held by anyone else.

Why CC0 suits data

Data is often combined from many sources. Each attribution or share-alike condition that travels with a source has to be honoured by everyone downstream, and the conditions multiply as datasets are merged. A large application that uses hundreds of sources can end up with a long list of attribution notices, and licences with incompatible conditions can make some combinations impossible. CC0 removes that friction entirely.

There is also the question of what rights exist. Facts are generally not protected by copyright, but databases can be protected by copyright in their selection or arrangement, and in the EU and UK by a separate database right. A publisher may not know exactly which rights apply to a given table. CC0 sidesteps the question: it waives whatever rights there are. Its legal code explicitly includes database rights under the EU Database Directive 96/9/EC and similar data extraction rights elsewhere.

The Open Definition's list of conformant licences places CC0 among its recommended licences, alongside the Open Data Commons PDDL, CC BY 4.0, ODC-By, CC BY-SA 4.0 and the ODbL. Data released under CC0 is therefore open in the sense used throughout this manual (see What is Open?).

Who uses CC0

CC0 is common for reference data meant to be built into other services. Wikidata, the structured-data sister project of Wikipedia, releases its data under CC0. Many cultural heritage institutions release catalogue metadata under CC0 so that it can be aggregated freely. Research data repositories also offer it as a default or recommended option for deposited datasets.

What CC0 does not do

• It does not cover rights you do not hold. If a dataset includes material owned by someone else, such as third-party map data or licensed imagery, CC0 cannot waive their rights. Check your sources before dedicating.
• It does not override privacy law. CC0 does not affect privacy or publicity rights. Open data should not contain personal data, as the manual notes in What Data are You Talking About?
• It does not cover trademarks or patents. A logo or product name in your data stays protected by trademark law.
• It cannot be taken back. You can stop publishing, but anyone who already has a CC0 copy keeps the freedom to use it.
• It provides no warranty. Users take the data as it is.

CC0 vs CC BY

The usual alternative is CC BY 4.0, which allows the same uses but requires attribution. The trade-off is straightforward:

	CC0 1.0	CC BY 4.0
Legal mechanism	Waiver plus fallback licence	Licence with conditions
Attribution required	No (can be requested)	Yes
Covers database rights	Yes	Yes (section 4)
Can be revoked	No	No, but rights end on breach (restored if cured within 30 days)
Friction when combining many sources	None	Attribution notices accumulate

Choose CC BY if credit matters for funding, reporting or trust in the source. Choose CC0 when maximum reuse matters more, especially for data designed to be combined with other data.

Asking for credit without requiring it

Many CC0 publishers still want to know their data is used and to be cited. They usually publish a short set of "community norms" or a suggested citation next to the data. These requests are not legally binding, but they are widely respected, particularly in research where citation is an academic norm in its own right. Keep such requests separate from the dedication itself so that the legal position stays clear.

How to apply CC0 to a dataset

1. Confirm you hold the rights (or that the data is not protected) and that the data contains no personal information.
2. State the dedication clearly on the dataset page, for example: "This dataset is dedicated to the public domain under CC0 1.0 Universal", with a link to the CC0 deed.
3. Record it in the metadata: in DCAT the dct:license property, in DCAT-US the license field, in schema.org the license property, using the URL https://creativecommons.org/publicdomain/zero/1.0/. See metadata standards for open data.
4. Include a short licence note inside downloads (a README file), so the terms travel with copies.
5. Optionally add a suggested citation and community norms.

The manual's chapter Apply an Open License explains why applying a licence or dedication is necessary even when the existence of rights is uncertain: it gives users clarity.

CC0 and the PDDL

Open Data Commons publishes a similar tool, the Public Domain Dedication and License (PDDL), written specifically for databases. Both are on the Open Definition's recommended list. CC0 is more widely recognised outside the open data community; the PDDL's text is tailored to databases. Either achieves the same practical result.

Frequently asked questions

Is CC0 a licence?

Strictly, CC0 is a public domain dedication: the rights holder waives copyright and related rights. Because a waiver may not be valid everywhere, CC0 includes a fallback licence granting the same freedoms, so in practice it works like the most permissive licence available.

Does CC0 cover database rights?

Yes. The CC0 1.0 legal code lists database rights under the EU Database Directive 96/9/EC among the rights it waives, along with similar data extraction rights elsewhere.

Can I withdraw CC0 later?

No. The CC0 legal code describes the waiver and the fallback licence as irrevocable. You can stop distributing the data, but copies already released under CC0 remain free to use.

Do users have to credit me if I use CC0?

Not legally. You can ask for a citation as a matter of good practice or community norms, but CC0 itself imposes no attribution condition.

Does CC0 let people use personal data or trademarks in the dataset?

No. CC0 does not waive patent or trademark rights, and it does not affect other people's rights such as privacy or publicity. Personal data should not be in an open dataset in the first place.

More practical guides

• CSV vs JSON
• Metadata standards (DCAT)
• FAIR data principles
• CC BY 4.0 for data
• Open Database License (ODbL)
• Database rights