Practice area
Copyright: protecting and enforcing intellectual property
Copyright governs the protection of intellectual property. It secures authors' rights in their works and regulates their use by third parties.
Its principal aims are protecting intellectual property, encouraging creative work and spreading knowledge and culture. We help you protect your intellectual property and pursue any infringements.
What copyright secures
- Recognition of authorshipThe right to be named as the author of the work, and the decision whether and how that naming happens.
- PublicationThe decision whether, when and in what form a work is made available to the public.
- ExploitationThe rights of reproduction, distribution and communication to the public, and with them the basis of every licence agreement.
- Appropriate remunerationThe claim to remuneration proportionate to the use granted, even where the contract provides otherwise.
Where we come in
Day to day the question is rarely whether a work is protected, but who holds which rights and who may use them to what extent. We draft and review licence and usage agreements, clarify the rights position for software, text, image and video, and pursue infringements of your rights out of court and before the courts. Conversely we defend you when a third party asserts claims against you.
From AI training to infringement
What an AI model may be trained on
Training means reproducing third-party works. The text and data mining exception permits it, but only where access was lawful and the rightholder has not reserved the use in machine-readable form. Disputes turn almost always on that reservation, and the burden of proof lies with whoever relies on the exception.
Content generated by AI
A purely machine-generated result is not a work, because the personal intellectual creation is missing. That does not leave it unprotected so much as initially free, and the interesting questions lie in the human reworking and in the material used as input.
Rights in the output of an AI system
Copyright, contract and trade secret law each answer differently who owns the output. The provider terms govern what you may do with it, and they rarely say the same thing as the statute.
Knowing your own portfolio
Sorting out who owns what once a dispute has started is too late. Copyright arises in a business continuously and incidentally, in software, text, images and video. The IP scan records the portfolio and exposes the gaps.
When rights are infringed
Whether there is an infringement is one question, which route is worth taking is another. Evidence, urgency and cost risk decide between a warning letter, an interim injunction and proceedings. Conversely we defend you when a third party asserts claims against you.
Unclear who owns the rights?
We look at your contracts and tell you what you may actually use.
Clarify the rights positionTopics in this practice area
- AI for developersAI in software development: licence risks in generated code, trade secrets in prompts, liability and the role under the AI Act.
- AI-generated contentCopyright in AI output: does protection arise, who holds the rights, and when does an output infringe someone else rights?
- Choosing an AI toolThe questions to ask before procuring an AI tool: role, data, rights, records and the points that belong in the contract.
- Consortium and grant agreementsHorizon Europe: what has to be settled before acceding to the grant agreement, how far the ethics requirements reach and where the AI Act applies.
- Inputs as training dataAre our inputs used for training? Confidentiality, trade secrets and how to draft the contract with AI providers.
- Passing on rights in AI outputPassing AI output to clients: which rights can be granted, which warranties are risky, and when labelling is required.
- Receiving a warning letterWarning letters under trademark, copyright or competition law: assessing the claim, modified undertakings and fending off unfounded demands.
- Rights in inputs and outputsWhat AI provider terms say about rights in inputs and outputs, and what an indemnity is actually worth.
- Text and data miningWhat an AI model may be trained on: the exceptions in sections 44b and 60d UrhG, the machine-readable reservation of rights, and the burden of proof.
- Transparency and labelling dutiesArticle 50 AI Act: people must learn when they talk to a chatbot, and generated content needs marking. Deadlines and implementation.
- Transparency code of practiceThe Code of Practice on Transparency of 10 June 2026: what it requires, what signing it achieves, and how to show compliance without it.
Frequently asked questions
Do we have to register our work somewhere?
No. Copyright protection arises with the creation of the work and requires no registration. In practice the question is therefore rarely whether a right exists, but who holds it and who may use it. Both are settled by the contract, not by a register.
A designer delivered a graphic to us. Does that make the rights ours?
Not automatically. Under German law copyright itself stays with the author and cannot be transferred. What is transferred are rights of use, and only to the extent the contract expressly provides. Where there is no provision, the scope follows the purpose of the contract, which regularly turns out narrower than expected.
May we use images and texts from AI tools commercially?
That depends on the terms of the service and on the training material. Two questions have to be kept apart: whether any protected right arises in the output at all, and whether the output infringes someone else's rights. We assess both for your actual use.
Who owns works created by our staff?
Copyright itself stays with the person who created the work, and only rights of use can be transferred. For computer programs the economic rights pass to the employer by operation of law where the program was created in the performance of duties. For everything else the employment contract decides, and where it is silent, the purpose of the employment.
May we use third-party images with a credit?
A credit on its own permits nothing. The quotation right requires a quotation purpose, meaning a genuine engagement with the work taken over. An image that merely decorates an article is not a quotation but a use that needs a licence.
What may an AI model be trained on?
Through the text and data mining exception, using lawfully accessible works, so long as the rightholder has not reserved the use in machine-readable form. For scientific research the exception reaches further and a reservation is ineffective. The burden of proving the conditions lies with whoever relies on the exception.
Our agency worked with AI. What does that mean for us?
You should read the warranties in the contract for whether they cover AI output. Nobody can grant more rights than they hold, and a purely machine-generated result regularly attracts no copyright. The indemnity for third-party claims should therefore expressly cover AI output as well.