The European Commission has reopened one of the most difficult EU AI copyright questions: what should happen when AI systems use copyright-protected work for training, generation and imitation. A new EU copyright and AI consultation, published on 29 September 2026, asks whether Europe’s existing framework needs additional measures as technology and creative markets change.
The exercise is not a new law and it does not predetermine what the Commission will propose. It is a targeted consultation, open until 3 November 2026, gathering evidence from rights holders, generative AI providers, intermediaries, researchers, consumer organisations, national authorities and other affected groups.
EU AI copyright goes beyond training data
The Commission says the consultation focuses on four broad areas: copyright-protected content used in artificial intelligence, online piracy of time-sensitive content such as live events, equitable remuneration for music performers and producers, and copyright issues in scientific research.
The consultation questionnaire itself shows how detailed the AI section has become. It separates questions about generative AI, licensing and enforcement, and AI-generated imitations of performers. That structure matters because the public argument often compresses several distinct disputes into a single phrase such as “AI copyright”.
Training is one problem. Licensing is another. Enforcement is another. Then there is the question of outputs that imitate a recognisable performer without necessarily reproducing a specific protected recording or image. Each can involve different rights, evidence and remedies.
Creators and AI companies are arguing over different bottlenecks
Rights holders have spent years arguing that AI developers need clearer permission and compensation when protected work contributes to model training. Developers, meanwhile, have warned that licensing systems can become impractical if rights information is fragmented, machine-readable signals are inconsistent or vast datasets contain material whose ownership is difficult to establish.
The consultation does not resolve those positions. Its job is to collect evidence about whether the current European framework works in practice and where additional measures might be justified. That makes the process less dramatic than a new regulation, but potentially more important to the shape of any future proposal.
The dispute is already visible in courtrooms and commercial negotiations. LiveAIWire has reported on the music industry lawsuit involving Anthropic and major publishers, one example of rights holders testing existing law against the way generative models are developed and used. Europe’s consultation is examining the policy environment before any specific legislative choice is made.
Imitation has become its own copyright-era problem
One of the more revealing parts of the questionnaire concerns AI-generated imitations of performers. Generative systems can produce voices, appearances and styles that evoke a person even when the output is not a literal copy of a protected work. That puts copyright alongside neighbouring rights, personality rights, consumer protection and deepfake rules.
LiveAIWire has also followed new pressure on platforms over intimate deepfakes. That is a different legal issue, but both debates show why synthetic media is difficult to govern through one body of law. A system can create something novel in a technical sense while still affecting the economic or personal interests of a real creator or performer.
For musicians and other performers, remuneration is part of the same wider argument about who captures value when technology changes distribution and production. The Commission is separately asking about the application of equitable remuneration rights to recorded music, which means the consultation is not solely an AI exercise even though generative AI is one of its most prominent themes.
Nothing changes immediately, but the evidence could shape what comes next
Businesses should therefore avoid reading the consultation as a fresh compliance deadline. Existing EU copyright rules remain in force. What is changing is the policy record: the Commission is formally collecting information on where stakeholders believe current rules are unclear, hard to enforce or poorly matched to new technology.
The strongest submissions are likely to matter most when they can show concrete problems rather than broad positions. For rights holders that could mean evidence about licensing, attribution, market substitution or enforcement costs. For AI providers it could mean evidence about technical feasibility, rights-reservation mechanisms and the practical consequences of different licensing models.
The immediate story is therefore not that Europe has chosen a side in the AI copyright fight. It has chosen to reopen the file with a narrower set of questions and a live deadline. Once the consultation closes on 3 November, the harder stage begins: deciding whether the evidence supports new measures, better enforcement of existing ones, or a mixture of both.
The timing also matters for AI companies operating across Europe. A consultation does not create a new duty, but it signals which questions regulators are actively testing. Companies that train or deploy generative systems may want clearer records of where content comes from, which rights signals they can detect, how licensing is handled and how complaints about outputs are investigated. Those practices can be useful even if the eventual policy response is limited.
For creators, the same period is an opportunity to move from principle to evidence. Showing how licensing mechanisms fail, where attribution breaks down or how synthetic imitation affects a market is more useful to a policy process than simply repeating that AI is either theft or innovation. The consultation is designed to gather that kind of detail before the Commission decides whether another intervention is justified.
About the Author
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, cybersecurity and the social impact of emerging technology. LiveAIWire is an independent, human-led technology publication using AI-assisted research, editorial production and original AI-assisted editorial illustrations under his direction.