Copyright in life sciences: what is protected and where the limits are

Key Takeaways

• Copyright does not protect scientific findings, ideas, methods or data as such. It protects the concrete form in which they are expressed, for example in text, software code, graphics or technical illustrations, provided there is sufficient originality and human creative input. 

• Not every work product meets the threshold for copyright protection. The more a document or output is shaped by scientific standards, regulatory requirements or technical necessity, the harder it is to establish an individual creative contribution. 

• Copyright arises automatically, but it still needs to be secured in practice. Unlike patents or trademarks, copyright does not need to be filed or registered. In a dispute, however, authorship, date of creation and usage rights must be provable. 

• Under German copyright law, the creator generally remains the author. Companies therefore need to ensure that they receive the necessary usage rights from employees, service providers, software developers, agencies and research partners. 

• AI makes the analysis more difficult. Pure AI-generated outputs may create protection gaps. Clear rules on input data, outputs, usage rights, confidentiality and documentation are therefore essential.

This article was co-authored with Daniel Schuppmann, Senior Associate in NEUWERK’s Life Sciences team.

Copyright as part of intellectual property

When people talk about innovation, research and competitive advantage, the term intellectual property is never far away. It refers to rights in certain intangible assets, meaning legally protected outputs and identifiers that are not physical objects.

Intellectual property is not one single legal category. It includes different rights, such as patents, utility models, trademarks, designs, trade secrets and copyright. In life sciences, discussions around intellectual property often focus on patents. That is understandable. Patents frequently protect the technical core of an innovation, such as an active ingredient, a manufacturing process or a medical use.

Copyright may appear less important at first glance. But that impression is misleading. While patents may protect technical research results, copyright often accompanies many steps along the way: scientific articles, study reports, database structures, software, technical drawings, training materials and molecular visualisations may all be relevant from a copyright perspective.

The key limitation is this: copyright does not protect the research result as such. It protects the concrete form in which content is expressed.

When does copyright protection arise?

Under German copyright law, the starting point is the concept of a “work”. Only outputs that qualify as works are protected. The law lists several categories by way of example, including literary works, written works, computer programs and scientific or technical illustrations, such as drawings, plans, maps, sketches, tables and three-dimensional representations.

This means that many outputs created in research and development can, in principle, fall within the scope of copyright law. But the decisive question is not whether a document is scientifically important or commercially valuable. The question is whether it reflects a personal intellectual creation.

In practical terms, that means human, individual creative input. This also distinguishes copyright-protected works from outputs created without specific human creative influence by machines or AI systems. For AI-generated content, it is therefore important to assess whether a human being has shaped the protectable elements of the result or whether the output is merely automated.

The creative contribution must also take a sufficiently concrete form. A mere idea, scientific concept or biological mechanism is not protected by copyright. Protection may arise only in the concrete expression, such as a text, graphic, software structure or visualisation.

There also needs to be a minimum level of individuality. The threshold is not high, but it exists. In scientific articles, individuality may lie in the selection, structure and language used to present the content. The scientific finding itself is not protected.

Copyright protection is often lacking where the output consists only of factual information or highly standardised presentation. Typical examples include:

• raw data and measurement results, such as raw clinical study data 

• simple schematic tables or standardised diagrams 

• short descriptive texts without individual character 

• highly formalised documents, such as standard SOPs, where there is no meaningful room for creative choices 

This is a recurring tension in life sciences. The more an output is dictated by scientific standards, regulatory requirements or technical necessity, the harder it is to identify an individual creative contribution.

If the requirements are met, copyright protection arises automatically at the moment of creation. Unlike patent protection, it does not require filing or registration. There is also no examination by an authority. Copyright generally lasts for the lifetime of the author and expires 70 years after the death of the author, or the last surviving co-author. Usage rights in protected works can therefore be economically relevant assets.

Which materials may be protected in practice?

The abstract requirements become easier to understand when applied to typical outputs from research, development and commercialisation.

Scientific articles and study reports are often protected as written works. Copyright does not protect the scientific findings themselves. It protects the specific language, structure and presentation of the content. Even shorter elements, such as abstracts, may be protected if the selection and compression of the content reflect an individual contribution.

Graphics showing biological processes may also be protected. Complex scientific illustrations often involve individual choices around structure, perspective, selection, colour and explanatory design. The less the presentation is fully dictated by scientific convention, the more likely protection becomes.

Software and analysis tools are particularly relevant in practice. Copyright mainly protects the concrete source code and, depending on the design, certain structural elements of the program. It does not automatically protect the function, scientific method, algorithm as such or technical idea behind the software. This distinction matters for sequence analysis tools, data-processing systems, digital health applications and AI-based analytics tools.

Databases and data collections may be protected on different levels. Individual data points or facts are generally not protected by copyright. However, the specific selection or arrangement of a database may be protected as a database work if it is creative. Separately, a database right may exist if there has been substantial investment in obtaining, verifying or presenting the data.

Molecular models and 3D visualisations may be protected as scientific or technical representations. This is particularly relevant where the output is not merely a standard, automatically generated representation of facts, but an individual visualisation created within a certain design space.

Training materials, SOPs and study documents may be protected if they go beyond purely standardised or functional instructions. Individual structure, explanatory text, graphics and didactic choices may be sufficient. The more a document is determined by regulatory or technical requirements, the more likely it is that practical protection will rely less on copyright and more on confidentiality, contractual control and, where appropriate, trade secret protection.

What does AI mean for copyright protection?

AI does not make copyright analysis impossible, but it changes the questions.

Under German copyright law, protection generally requires human creative input. Pure AI-generated outputs without human creative contribution are therefore usually not protected as copyright works. Protection may be more plausible where a human being shapes the protectable elements through selection, editing, arrangement or post-processing and uses the AI merely as a tool.

For life sciences companies, this is practically relevant where AI is used for texts, images, software development, data analysis, molecule concepts, documentation or scientific visualisations. The question is not only whether the output is protected by copyright. It is at least as important to ask what rights exist in the input data, whether confidential information may be entered into an AI system, whether outputs can be used exclusively and whether the provider may use prompts, inputs or outputs for training purposes.

In practice, protection will often not rest on copyright alone. Contracts, trade secrets, access controls, documentation and clear internal AI policies will usually be at least as important.

How does copyright work in a company context?

For companies, the relevant question is not only whether a work is protected by copyright. It is also who is allowed to use it.

The starting point under German copyright law is that the author is the natural person who created the work. That may be the scientist who wrote a technical text, the designer who created a molecular visualisation or the developer who wrote software code. Authorship arises with the creation of the work and generally cannot be transferred by contract.

What can be transferred or granted are usage rights. These are the rights to use, reproduce, publish, edit or commercially exploit a work in a particular way.

In practical terms, this means that a company under German law does not usually become the author of a work merely because the work was created in a business context or paid for by the company. The company needs the relevant usage rights. For employees, those rights may follow from the employment contract, the purpose of the employment relationship or specific statutory provisions. For computer programs, German copyright law contains a special rule in favour of the employer where the program was created in the performance of the employee’s duties or under the employer’s instructions.

Contractual arrangements become even more important with external partners. Companies working with agencies, service providers, software developers, scientific advisers, CROs, universities or research partners should expressly define which usage rights are granted. The issue is not only whether content may be used at all. Scope, duration, territory, exclusivity, editing rights, sublicensing, group-wide use and use in later products, databases, publications or transactions also matter.

Conflicts often arise where the scope of rights remains unclear. A company may have paid for a presentation, graphic, study or software tool. That does not automatically mean that all usage rights have been granted.

This approach is not universal. German copyright law is strongly shaped by the personal relationship between the author and the work. US copyright law is traditionally more exploitation-oriented. Copyrights can generally be transferred, and the concept of “work made for hire” plays an important role. This difference matters in international research collaborations, global software projects and group-wide IP arrangements. Clauses drafted from a US perspective may be ineffective or at least require interpretation under German law. Conversely, German-style drafting can create misunderstandings in international settings.

How is copyright enforced?

Unlike many other IP rights, copyright arises automatically when the work is created. That is often also the practical challenge. In a dispute, it must be possible to prove who created the work, when it was created and which rights exist.

Anyone who wants to enforce rights should document early:

• when the relevant content was created 

• who contributed to its creation 

• which versions exist 

• which usage rights were granted internally or externally 

• whether AI tools were used and what human contribution was made 

If content is used without authorisation, several remedies may be available. The main claim is often an injunction to stop ongoing or threatened infringement. Claims for damages, removal, information, destruction or recall of infringing copies may also be available.

Damages are often calculated by reference to a hypothetical licence fee. The question is what reasonable parties would have agreed as a licence fee for the specific use. This can be relevant in particular for software, scientific visualisations or commercially used technical content.

Evidence should be preserved early. If content is used without permission, screenshots, website captures, email trails and file versions should be documented quickly. Digital content can be changed or removed at short notice.

In many cases, enforcement does not begin with litigation. It starts with a cease-and-desist letter, a request to remove infringing content or a negotiation about retrospective licensing. Injunction claims may also be pursued through preliminary relief. Delay can be harmful if urgency is required. Companies should therefore not leave potential infringements unaddressed for too long if urgent measures may later be needed.

Practical steps for companies

For life sciences companies, copyright is most useful when it is documented and contractually secured. Practical steps include:

1. Identify relevant materials. Which texts, graphics, software, databases, training materials, study documents or visualisations have commercial value? 

2. Assess protectability realistically. Not every document is protected by copyright. Raw data, methods, ideas and highly standardised outputs often require other protection mechanisms. 

3. Document authorship and versions. Creation dates, contributors, versions and editing steps should remain traceable. 

4. Regulate usage rights by contract. Employment agreements, service agreements, software contracts, research collaborations and agency agreements should create clear rights chains. 

5. Control the use of AI. Input data, confidentiality, training use, output rights and human editing should be documented and regulated. 

6. Use confidentiality as an additional layer. For SOPs, study documents, databases and regulatory documentation, trade secret protection may be more important than copyright. 

7. Preserve evidence of infringement. Digital copies, screenshots, file versions and publication dates should be secured early. 

Conclusion

Copyright is not a substitute for patents in life sciences. It does not protect the active ingredient, the scientific finding, the technical method or raw data as such.

It is still practically important. It may protect the concrete form of scientific texts, software code, database structures, graphics, technical representations, training materials and visualisations. This layer of protection becomes increasingly relevant in data-driven and software-enabled life sciences models.

For companies, however, the decisive question is not only whether copyright arises. The more important question is whether the necessary usage rights are properly held by the company, whether AI use is controlled and whether complementary protection mechanisms, such as contracts, confidentiality and trade secrets, have been built into the overall IP strategy.