Copyrighting AI Generated Art

Evaluating copyright protection in light of emerging AI creations

Dr. Susanna Grech Deguara | 30 May 2023

Copywriting AI

Copyright and AI in a Nutshell 

Artificial Intelligence (AI) has taken the world by storm and is rapidly transforming various industries and activities. The realm of Intellectual Property (IP) is no exception. The impact of AI on IP is multifaceted, ranging from automated IP searches to the creation of AI-generated content. As AI continues to evolve and gain popularity, it raises complex legal and ethical questions around IP ownership and protection. 

Background to AI Generated Works    

Many AI systems produce creative works, such as music with Amper Music, translations with DeepL, poetry by Google’s Verse-by-Verse or paintings with ArtBreeder. The results are certainly artistic, musical or literary works; one cannot tell whether they have been created by a human or AI. However, this does not mean that they also benefit from copyright protection. 


Key Legal Issues of AI Generated Works  

  • Establishing ownership of AI created works
  • The extent of copyright protection awarded by law to AI created works

The Fundamentals  

AI is also being used to generate new forms of creative content, such as music, literature, and art. This raises questions around IP ownership and protection. The question which invariably arises is whether such works can also benefit from copyright protection.

Can AI Works Benefit from Copyright Protection?

Together with being an artistic work, which evidently, AI is able to create, for a work to be protected through copyright it must also constitute a concrete and original expression by an author. According to EU case law, works can enjoy copyright protection if they represent “the author’s own intellectual creation”1


This originality standard contains two elements. Firstly, the work must not have been copied, and secondly, that it presents an intellectual creation. While the former requirement can be fulfilled by AI, the latter cannot. Machines nowadays are able to create independent works that deviate sufficiently from the style they learned from and therefore can be considered novel2 . This notwithstanding,  the second element required is inherently linked to a human person, highlighting the pivotal role the author has in an anthropocentric system3.  

Free, Personal and Creative Choices

According to established CJEU case law4, a work benefits from copyright protection only if the author made free, personal and creative choices in the creation process, that reflect one’s own personality. Skill and ability that show economic investment alone do not suffice for the enjoyment of copyright protection5. This stance has not only been adopted in the EU, similar standards have been established by the US Supreme Court6

These human creative preferences vary fundamentally from how AI machines operate. Computers are able to carry out a solely mechanical, deterministic processes, on the basis of the information provided and programmed7. As things stand, AI is not able to generate an outcome without prior data about similar objects. This is fundamentally different from creative activity by humans, who create choreographies or music “by internal stimulus and without prior training;”. It is needless to say that AI will not be able to feel emotions or a need for self-expression, which is embodied in creativity. In conclusion, AI-systems as machines will not be able to make creative choices that bring the output they create in the realm of copyright protection.

Copyright for AI Assisted Output  

In contract to the above, works created though AI but which still hold a sufficient trace of human creativeness can enjoy copyright protection. The degree of human guidance in AI creations is yet to be determined.  

Human beings make creative decisions in the conception phase of works, regarding subject matter and plot, together with the medium, format or, if to be used, the AI tools. Such decisions as to which AI system to use or which data to be inputted. Additionally, in the redaction phase of a creative work, humans often apply corrections or edits and make final choices regarding the product before it is made available to the public. This process also applies to AI-assisted products. However, creativity is less likely to occur in the execution phase of an AI, particularly with a machine learning system for which the distance between input and output is large and cannot be fully preconceived or explained by a human.

Creative Choices and The Different Types of AI Systems

The creative choices, however, depend on the type of AI system at issue. Generally, there are three main types of systems. 

 Firstly, the step-by-step algorithm, that make us of concrete if-then rules without room for much deviation. The second type of AI systems is the rule-based algorithms that operate within margins provided by the programmer, leaving some room of operation to the AI system. Lastly, one can opt to make use of machine-learning algorithms that learn on the basis of input data and can generate an unknown style variation8.  Evolutionary algorithms will fall under this last category, where an AI system aims at optimizing samples according to predefined criteria by autonomously selecting, mutating and producing new samples coming closer to the requirements.  

What this Means for You 

What is clear is that the emergence of AI poses challenges for the traditional frameworks of IP protection. In response to these challenges, there is a growing need for legal frameworks that can accommodate the unique aspects of AI and its impact on IP. Some experts suggest that new IP rights may need to be created to address the unique challenges posed by AI-generated works, while others argue that existing IP laws can be adapted to accommodate the changing landscape. One might easily acquire or loose copyright protection over one’s own assets due to changes in legislation.  

How We Can Help  

In conclusion, AI is rapidly transforming the IP landscape, raising complex legal and ethical questions around ownership, protection, and liability. As AI continues to evolve, it is essential that legal frameworks keep pace with these changes to ensure that IP is protected and that the benefits of AI are realized. Our technology team lawyers work strive to keep abreast with the latest amendments together with helping clients understand how they can avail of intellectual property protection in light of emerging technologies. Contact us for more information on how you can protect your most valued assets.  



1 Case C-05/08 Infopaq International A/S v Danske Dagblades Forening (Infopaq) [2009] ECR I-06569, para. 45.

2 Martin Senftleben and Laurens Buijtelaar, ‘Robot creativity: an incentive-based neighbouring rights approach’ [2020] 42 EIPR 797, 799.

3 Case C-469/17 Funke Medien NRW GmbH v Federal Republic of Germany (Funke Medien) [2019] ECLI:EU:C:2018:870, Opinion of AG Szpunar, para. 60

4 Case C-683/17, Cofemel — Sociedade de Vestuário SA v G-Star Raw CV [2019] ECLI:EU:C:2019:721, para. 30; Case C-833/18, SI and Brompton Bicycle Ltd v Chedech / Get2Get [2019] ECLI:EU:C:2020:461, para. 230.

5 Case C-604/10, Football Dataco Ltd and others v. Yahoo! UK Ltd and others [2012] ECLI:EU:C:2012:115, para. 42.

6 Feist Publications, Inc. v Rural Telephone Service Co., 499 U.S. 340, 346 (1991).

7 Anna Shtefan, ‘Creativity and artificial intelligence: a view from the perspective of copyright’ [2021] 16 JIPLP 720.

8 Martin Senftleben and Laurens Buijtelaar, ‘Robot creativity: an incentive-based neighbouring rights approach’ [2020].

Request More Information

Please send me legal and other updates

Key Contacts

Dr Charlene Mifsud

Partner, Corporate & Commercial

+356 2205 6298

Related Practice Groups