Build1 publisher2 min readPublished
European Commission gathers evidence on copyrighted content in AI until 3 November 2026
European Commission is taking views until 3 November 2026 on how EU copyright law should treat AI use of protected content. Responses from AI developers and rights holders go into the evidence for rules the Commission has not yet decided to write.
The Engineer · Build desk
Drafted by a language model from the sources cited here and checked against its claim ledger before publication. How we use AISend a correction
What happened
- Through this consultation the Commission is gathering feedback on possible future measures and is not setting licensing obligations or technical requirements.
- The survey spans four policy areas that go beyond generative AI, including piracy enforcement on platforms and copyright rules for research on scientific materials.
- The Commission wants views from rights holders, general-purpose AI providers, AI developers, intermediaries, research bodies and consumer groups.
- The dev.to account expects the questions under review to shape future expectations on licensing, data use, transparency and enforcement across the single market.
Compiled by The EngineerSomething wrong?How this is made
Why it matters
- constraint Teams cannot yet plan EU product work against specific licensing or transparency terms, because none exist and the Commission has not committed to writing any.
- cost The per-dataset record of sources and licence terms that a useful response needs is likely the first thing any later transparency measure would ask for, so the effort carries over.
- exposure Rights holders answer the same survey, so an AI developer that stays out leaves its own operating constraints out of the evidence the Commission weighs.
The process has three steps, according to the dev.to account of the Commission's announcement.
1. A Call for Evidence, which the Commission ran earlier in 2026 [4]. 2. This targeted survey, whose closing date ends the public input stage [8]. 3. An evaluation of the survey responses together with the Call for Evidence. After that, the Commission decides whether to pursue future measures [8].
Step three is conditional. The account does not say what follows the evaluation or when a decision is due [16].
The source is specific about what a useful response contains. It should describe the activity at issue, the rights or data-use uncertainty involved, the practical effect on the business or sector, and the clarity or safeguard that would help [11]. It is a sound template. According to the source, it gives policymakers information they can use without overstating what the consultation can deliver [17].
For AI developers and providers, the first input is a map of the content and data types used in development, with their known licensing or usage terms [10]. The source also suggests that companies building AI capabilities look now at how they document data sources, permissions and contractual restrictions [15]. I'd expect filling in the survey form to be the quick part. The slow part is tracing which training set arrived under which terms, and whether anyone kept the contract.
The source expects content owners to approach the same questions from the other side. Some may want clearer ways to understand or control how their protected material is used in AI [14]. The source adds that neither position automatically determines what future EU policy will be [14].
The 3 November 2026 date only closes input [3]. The source advises businesses not to treat the consultation as a new legal deadline for their AI products [12].
What to watch
- A Commission summary of the survey and Call for Evidence responses that states whether it will pursue future measures.
- Any follow-on proposal that names specific transparency or licensing requirements for general-purpose AI model providers.