Home / Publications / The intellectual property dilemmas caused by AI-generated...

The intellectual property dilemmas caused by AI-generated works


With the rise of artificial intelligence (AI) in many sectors, questions pop up about the full extent of AI use and its influence on ownership and accountability. 

In healthcare, AI is used to support and advise doctors in establishing a diagnosis, assist radiologists in improving image interpretation, and improve drug discovery research, to name just a few of the current applications. 

However, with the rise of a new technologies, new issues arise too. If we take a look at the last example, an important question that comes up is: who owns what when AI does the research? To answer this question—and offer you advice on how to deal with this regarding your AI innovation—we’ll review patent and copyright law in this context.

Human intervention

Intellectual property is extremely valuable for your company. Not only can investments in innovation be recouped through the exploitation of intellectual property, you can also keep competitors at bay by enforcing these rights. In addition, the (re)structuring of intellectual property portfolios can provide strategic benefits.

Traditionally, intellectual property rights vest with the human creator or inventor. At the time when intellectual property laws came into being, it was clear who the inventor or creator was: the human(s) that created the work. This can still be the case with AI-generated works if AI is merely used as a tool and human intervention in such a work is clearly present. However, the latter is becoming increasingly difficult to establish as the technology progresses rapidly and AI is able to operate more independently. 


Under the European Patent Convention (EPC), you can’t patent software 'as such'. Case law shows that software is patentable when it can be presented as a new, non-obvious technical solution to a technical problem.

The European Patent Office (EPO) has amended its guidelines to provide more clarity on the circumstances under which inventions in the field of AI are patentable. According to these guidelines, AI-based inventions are patentable as long as the method used serves a technical purpose. For example, the use of a neural network in a heart-monitoring apparatus for the purpose of identifying irregular heartbeats makes a technical contribution and is thus, in principle, patentable.


For a work to be subject to copyright protection, originality is required and creative choices must be made. Most definitions of originality require a human author. Without human intervention in the creative progress of AI-generated works, copyright protection is not possible. For example, the Beijing Internet Court in China recently held that the production or the creation of a ‘work’ by a human is a prerequisite to be eligible for copyright protection. AI software could, therefore, not be considered an author.

The question then arises if there could be another solution to protect AI-generated works under copyright law. The general rule is that the one who creates the work is regarded as its author, not the one who physically makes the work.

However, an exception to this rule is employer's copyright, which is comparable to works made for hire in the United Kingdom and the United States. Employer’s copyright makes it possible for the employer to be regarded as the (notional) author if the employee created the work during the scope of his employment. An analogous application of works made for hire for AI is suggested in literature. This would give your company an incentive to continue investing in innovation.

What now?

A lot is still unclear in the field of intellectual property protection of AI-generated works. Since AI does not have legal personhood, the question whether it can be holder of any (intellectual property) rights can, for now, only be answered negatively. If no inventor or creator can be identified, the work will, in principle, fall into the public domain.

On the one hand, this means that information is easily accessible and new knowledge can be created unencumbered. On the other hand, investments can’t be recouped and valuable advantages over competitors will disappear. In the worst case, this could result in limited innovation and less investments in AI research.

Although the EPO has taken a first step towards examination procedures that are friendlier to AI-related inventions, it is still unclear what the future may hold regarding copyright for AI-generated works. Employer's copyright? New provisions regarding AI-generated works? Legal personhood for AI? Who knows what the near future will bring?


CMS is partner of Blue Tulip Awards, an award to accelerate startup innovations. This publication is exclusively created for the Health innovation theme. Please visit our Blue Tulip Awards page or the Health theme page for more information. 


Portrait of Yasar Celebi
Yasar Celebi