Back

Legal Challenges in Copyright Law in the Space of Generative AI Creativity

Written by Abdulsamad Jimoh and Adaeze P. Nwoba

One of the most fascinating examples of technological progression in the modern world is the emergence of generative AI. It is interesting to note that this development has raised significant copyright issues as to the true owner of the content, between human beings and the generative AI. This article offers an in-depth exploration of what generative AI entails and the current position of Nigerian copyright law to determine the best possible approach to address the challenges presented by this technology in the legal realm.

What is Generative AI?

Generative AI is a subgroup of Artificial Intelligence (AI). It is a sophisticated machine that has been technically trained with the capability of learning complicated algorithms to create creative works like written content, music, and images. One of the famous AI tools for generating text is ChatGPT, which was developed by OpenAI in 2020. It is highly sophisticated in generating text as it can write articles, stories, essays, poems, and also provide prompt answers to diverse complex questions directed to it.

The generative AI is also utilised in image and audio production. Some of the notable AI tools for generating images include Stable Diffusion, DALL·E 2, and Midjourney, which can create stunning visuals in diverse styles such as aged photography, pointillism, and pencil drawings. Similarly, AI tools like Aiva and Amper Music are revolutionising the music industry by generating original compositions.

These AI tools can appear to be magical because of their ability to create astonishing and realistic outputs for the users, but the truth is that they are far from that as their programme is grounded in complex algorithms and immense training on vast datasets. The systems employ unsupervised machine learning models that allow them to detect content based on patterns in data without human interference.

Impact of Generative AI on Copyright

The emergence of Generative AI has obviously posed serious challenges to the traditional notions of authorship and ownership. This development has brought new dimension to the realm of creativity, leading to a complete dilemma as to the real owner of the content, between human beings and the generative AI.

One of the challenges is the fact that the AI tools make use of copyrighted materials to create content. This has raised questions as to whether the use of these materials constitutes copyright infringement. Recently, legal actions have been taken by original owners against AI companies. One of the popular instances include that of American author and comedian, Sarah Silverman, along with two other authors who instituted a lawsuit against OpenAI and Meta for alleged copyright infringement, claiming that the companies’ AI models used their work in their training data without permission.

A further example is the US case of Andersen v. Stability AI and others, where three artists brought a class action against some generative AI platforms on the ground that the developers, without a license, utilised their original works to train their AI in their styles and made it possible for users to generate works that may be insufficiently transformative from their existing protected works, and, as a result, would be unauthorised derivative works. However, the court made it clear that the plaintiffs must allege that the outputs are “substantially similar” to their original artwork to substantiate their copyright claims.

The second challenge goes to the ownership of AI-generated content. This aspect has caused serious concern about whether AI-generated content can be copyrighted and who should be the rightful copyright owner. The relevant case in this regard is that of Kris Kashtanova, who registered a copyright for a graphic novel, Zarya of the Dawn, illustrated with images that Midjourney generated in response to text inputs in September 2022. In October 2022, the US Copyright Office initiated cancellation proceedings, stating that Kashtanova had not revealed the use of AI. Kashtanova responded by arguing that the images were made via a creative, iterative process. In February 2023, the Copyright Office decided that the photos were not copyrightable, deciding that Midjourney, rather than Kashtanova, authored the visual material. In March 2023, the Copyright Office released guidance declaring that when AI determines the expressive elements of its output, the generated material is not the product of human authorship. While the book’s text is protected by copyright because they are Kashtanova’s creative work, the images are not as they are AI-generated content.

The direction taken in Kashtanova’s case reaffirmed the US position on copyright that human creativity is necessary for ownership to exist.

This is slightly different in the UK, where both human creative works and content generated with limited human interference are recognised. The UK Copyright, Designs, and Patents Act of 1988 brings the latter to the picture through its definition of computer-generated works as those works generated by computers in circumstances where no human author exists. By implication, this definition encompasses the modern generative AI tools. The Act, however, recognises that the works created by human authors will be copyright-protected for 70 years plus the author’s lifetime, while the ones generated by computer will be granted protection for 50 years from the time the author generated them.

Copyright Law in Nigeria

Currently, in Nigeria, copyright is regulated by the Copyright Act, 2022, which became effective on March 17, 2023, repealing the previous Copyright Act, Cap C28, LFN, 2004.

The principal objectives of the new law include: the protection of the rights of authors and ensuring just rewards and recognition for their intellectual efforts, and the provision of adequate limitations and exceptions to guarantee access to creative works.

In essence, the law exists to safeguard the creator of an original work that has been expressed in a particular way, preventing unauthorised individuals from using, copying, or modifying it.

According to Section 2(1) of Copyright Act, 2022, the following works shall be eligible for copyright:

  • Literary works
  • Musical works
  • Artistic works
  • Audiovisual works
  • Sound recordings
  • Broadcasts

Section 2(2) of the Act clarifies that literary, musical, or artistic work can only be eligible for copyright if two conditions are met:

(a) The work must possess original character, and

(b) It must be fixed in a medium of expression.

The provisions of the law centre around two key elements: originality and fixation.

  • Originality:

This denotes that the work must be an outcome of independent creative expression. It should stand as a unique creation.

  • Fixation:

This requires that the work must be stored in a tangible form of expression. Fixation means the work should be enduring enough to be seen, copied, or shared for a considerable period.

Section 19 of Copyright Act varies the duration of copyright as follows:

(a) For literary, musical, or artistic works (excluding photographs), it lasts 70 years after the author’s death.

(b) For government, state, and international bodies, it is 50 years after the year of public availability or creation if not shared publicly.

(c) Audiovisual works and photographs enjoy 50 years after public availability with author’s consent or creation if not shared.

(d) Sound recordings have 50 years after public release with author’s consent or creation if not shared.

(e) Broadcasts retain rights for 50 years after their initial transmission.

In cases of joint authorship, the reference is to the last surviving author’s death.

AI and Copyright in Nigeria

Nigeria has played a pioneering role in AI progression on the African continent, having established the National Centre for AI and Robotics (NCAIR), a government institution responsible for fostering AI research and improvement in the country.

Although the country lacks a structured national AI policy, the National Information Technology Development Agency (NITDA) and NCAIR and other vested stakeholders spearhead advancement in this domain. Significantly, diverse government institutions and departments, including the Nigeria Communications Commission, the National Board for Technology Incubation, the National Office for Technology Acquisition and Promotion, the Ministry of Communication and Digital Economy, and the Federal Ministry of Science and Technology, are actively involved in the adoption and exploration of AI and other emerging technologies.

In December 2023, the Attorney General and Minister of Justice, Lateef Fagbemi SAN, sought stakeholders’ collaboration to improve the copyright system to elevate Nigeria’s creative economy. He noted that copyright development is currently receiving considerable attention across the globe, and Nigeria must not be left behind.

Presently, the Nigerian Copyright Act only recognises human authorship and not that of a machine. This implies that the lawmakers should have envisaged that AI and machines would make inventions in the future, which is now. In essence, the position of the law on copyright to generative AI may be imprecise in Nigeria. However, the current law can still be extended to it where it infringes upon an existing copyright.

Therefore, when the use of AI violates the provisions of the Copyright Act, particularly Section 36, it would be deemed an infringement of copyright, and the violator may be liable before the court.

Conclusion

In conclusion, AI is an important facet of technology that is yet to be fully regulated in many parts of the world, including Nigeria. There is a current discourse regarding the scrupulous equilibrium between regulatory measures and the pursuit of innovation, which is crucial given the vibrant nature of technology. It is highly essential that we continue to have these discussions to determine the best course of action. Addressing the legal consequences of AI is indispensable, and it requires a multifaceted approach that includes explicit regulatory structures, ethical standards, sector norms, and cooperation between legal experts, policymakers, and technologists.

 

References

Leave a Reply

Your email address will not be published. Required fields are marked *