Glossary · Rule layers and human rights
Co-regulation (AI)
Also known as: Co-regulation, Regulated self-regulation
German: Ko-Regulierung
Co-regulation is a form of regulation in which a law sets the objectives and obligations and recognizes codes, standards or other instruments drawn up by or with the regulated parties as a way of meeting them; the instruments are voluntary, but they have a defined legal effect.
- AI regulation
- EU
- USA
- Canada
- China
In one sentence
Co-regulation: the law sets the goals, recognized codes and standards show how to meet them. The layer between AI law and provider rules.
Example
A provider of a general-purpose AI model that adheres to the GPAI Code of Practice can rely on it to demonstrate compliance with its obligations under the AI Act; the code itself is not law.
Explained in context
Context cards connect this term with others to answer one question. Also in British English and German.
How it applies
- Between law and company rules: Co-regulation is the layer between binding law and the rules a provider sets for itself. Its instruments are voluntary to adopt, but the law says what adopting them achieves.
- EU: Under the AI Act, harmonized standards give a presumption of conformity once their references are published in the Official Journal (Article 40); the GPAI Code of Practice (July 10, 2025) lets providers of general-purpose AI models demonstrate compliance (Article 56); codes of practice support the transparency duties of Article 50.
- USA: The NIST AI Risk Management Framework is voluntary; it acquires effect where a law, a contract or a procurement rule refers to it.
- Canada: The Voluntary Code of Conduct on Generative AI (September 2023) is self-commitment without a statutory link.
- China: National standards (for example GB 45438-2025 on labeling) give technical form to the rules of the Cyberspace Administration.
- Caution: Signing a code or applying a standard is not conformity in itself, and an uncited standard confers no presumption.