The AI Copyright Dilemma

 


The AI Copyright Dilemma

Dareem Scipio*


Since the public release of artificial intelligence technologies such as OpenAI last year, the United States Copyright Office (USCO) has reported a significant increase in persons seeking copyright protection for works generated by these technologies. This development has raised questions about whether AI generated works meet the long-established criteria necessary for copyright protection, particularly the requirement of human authorship.[1]

 

On a strict interpretation of the current laws relating to authorship, any work solely generated by AI generated technologies will be incapable of obtaining copyright protection. The challenge involves those cases where there is a mixture of human involvement or authorship and mechanical ‘designer-ship’ by AI technologies. In such cases, the primary issue becomes whether the level of human involvement is such that the resulting work constitutes an original work of human authorship. Even then a further question arises as to whether the whole work will be protected or only the aspects that have satisfied the requirement of original human authorship.

 

THE CASE OF KASHTANOVA; THE UNITED STATES POSITION


There is one example in the United States that provides some direction on how applications for copyright protection will be assessed in light of the emerging technologies. Kris Kashtanova was able to successfully register her comic book Zarya of the Dawn which was produced using AI- generated images, a fact that was not known to the USCO at the time the grant was made. However, in a not surprising turn of events, the USCO reversed its original registration and instead allowed protection for only part of the book when it became known to them that the images in the book were AI generated. Kashtanova was only granted protection for the words that accompanied the images, that is the parts that were her original literary creation.[2]

 

As a precaution for the future, the USCO also stipulated that persons submitting works for registration in the future must declare if AI was used in any part of the work. 

 

 

THE UNITED KINGDOM LEGAL POSITION


The legal systems of some countries such as the United Kingdom can be viewed as keeping up with the times. The challenges and ambiguity that now confronts the legal system in the US and other countries that follow similar legal systems, are addressed somewhat by a unique provision which makes the creator of an AI generated program the author of any work(s) created by that program[3]. Section 9(3) of the UK Copyright, Designs and Patents Act (CDPA) states 'In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken.'

 

However, even this provision leaves much to be answered. The question of who exactly is to be considered as “the person by whom the arrangements necessary for the creation of the work are undertaken” becomes an issue. Is it the person who has created the AI system itself? Or is it the ordinary user who uses the available system to produce a piece of work for him/herself? Adopting either approach highlights some issues. The former approach appears to be entirely at odds with the purpose of copyright protection and is impractical at best. It means that the developer of an AI software or program will potentially be deemed the author of any content. On the other hand, adopting the latter approach will mean, at least in cases where very little human involvement is necessary, that the law will be protecting works that are not original creations/does not involve human authorship, which also goes against the purpose of copyright protection.

 

WHAT IS THE LIKELY IMPACT IF AI WORKS CAN’T OBTAIN PROTECTION?


The likely consequence is that AI generated works may be used by anyone without consequences.[4] However, as these technologies continue to develop and become more mainstream, this may lead to challenges and conflicts. In that regard, copyright law must find a way to keep up with these rapid technological developments. While the UK’s legal approach highlighted above is a step in the right direction, it does not fully address the issue in a practical way keeping in mind the purpose behind copyright protection. The approach taken by the US, while more practical and keeping within the purpose of copyright protection, may prove to be a challenge as these technologies improve and it becomes more difficult to differentiate between AI generated works and works of original human authorship. Indeed, relying on human beings to disclose whether AI was used in the creation of a work is not the best authorship verification test.

 

CARIBBEAN IMPACT?


Since the copyright systems of the majority of Caribbean countries largely follow the traditional UK approach, and are increasingly being influenced by US copyright concepts, it remains to be seen how the issue of AI created works will be handled in the region.



* Author can be contacted at studentscipio@gmail.com

[1] Compendium II Copyright Office Practices s 202.02 (b)-The term "authorship" implies that, for a work to be copyrightable, it must owe its origin to a human being; s 503.03 (a) - Works produced by mechanical processes or random selection without any contribution by a human author are not registrable. Also see, Trinidad and Tobago Copyright Act, section 3; “author” is the natural person who has created the work.

[2] Shanti Ecalante-De Mattei, ‘US Copyright Office: AI works are not eligible for copyright’ ARTnews (2023).

[3] Andres Guadamuz, ‘Artificial Intelligence and Copyright’ WIPO Magazine (2017).

[4] ibid.







Comments

Popular posts from this blog

Angostura Queen’s Park Swizzle: More Than an Ounce of IP in Your Favourite Cocktail

Securing Copyright In Soca Music

Licks for 'Lick'

Broadcasting Bacchanal: Watching Football and Copyright Law

The Cost of Copyright Infringement