18 October 2023

Recent developments in the field of artificial intelligence (AI) have brought to the forefront a critical legal issue: the protection of intellectual property rights in the digital era. In the United Kingdom, this concern has garnered significant attention, with publishers urging the Prime Minister to safeguard the rights of authors and content creators.

The US Lawsuit Against OpenAI 

One pivotal event in this discourse is the lawsuit that unfolded in June 2023, where two US authors filed a class action against OpenAI in a San Francisco federal court. Their claim asserts that OpenAI misappropriated their works for the training of ChatGPT, a widely used generative AI system. This lawsuit serves as a striking example of the tension between AI development and the protection of intellectual property rights.

The Publishers Association's Plea

In response to these concerns, the Publishers Association, representing both digital and print book publishers, research journals, and educational content providers, penned a letter to the UK Prime Minister, Rishi Sunak. This letter urged the government to emphasise, during an upcoming summit on AI, the paramount importance of respecting intellectual property law when AI systems utilise content generated by the UK's creative industry.

Within the UK, the government initially considered allowing AI developers unrestricted access to copyrighted books and music for training AI models. However, this exemption, proposed by the Intellectual Property Office in June 2022, has since been retracted by ministers. A parliamentary report, published recently, criticised this initial proposal, emphasising a "clear lack of understanding of the needs of the UK's creative industries."

Copyright of AI-Generated content

In the United Kingdom, the Copyright, Designs and Patents Act of 1988 has provided a legal definition for computer-generated work, categorising it as content "generated by computer in circumstances such that there are no human authors of the work." This legal stipulation implies that content produced by artificial intelligence (AI) systems can indeed be safeguarded by copyright. Nevertheless, the provenance of responses generated by AI chatbots, and whether they might incorporate copyrighted materials, poses a unique challenge.

Two Key Questions Arise:

1. *The Use of Original Third-Party Content: * One pivotal query pertains to whether AI systems like ChatGPT should be permitted to employ original content generated by external parties to formulate their responses.

2. *Authorship of AI-Generated Content: * The second query delves into the issue of authorship. Is authorship exclusive to humans, or can AI systems be regarded as authors, especially when they produce creative outputs?

The technology underlying ChatGPT is founded upon a Large Language Model (LLM) architecture. To enhance its capabilities, these models are exposed to vast datasets, encompassing extensive web content and literary works. Presently, the UK allows AI developers to engage in text and data mining (TDM), albeit strictly for non-commercial purposes. OpenAI's terms of use grant users "all its right, title, and interest in the output."

However, the responsibility to ensure that the usage of this content complies with relevant laws lies with the users. Furthermore, these terms and conditions are subject to alteration and lack the stability and legal weight of traditional copyright. Consequently, the only viable solution lies in the clarification of laws and policies. Otherwise, organisations may need to resort to individual legal actions to assert their ownership of content utilised by AI. The absence of government intervention could potentially lead to the widespread use of copyrighted materials without the original author's consent.

Turning to the second question: Who can claim copyright over AI-generated content? Content created by AI will not be owned by the AI itself. Indeed, under European (and US) law AI cannot own copyright, as it cannot be recognised as an author and does not have the legal personality which is a pre-requisite for owning (intangible) assets.

Copyright law is grounded in the fundamental principle that only content created by human beings can be eligible for protection. While the algorithms underpinning ChatGPT were developed by OpenAI, copyright protection over chatbot responses may not be as straightforward.

Conclusion

The intersection of AI and intellectual property rights presents complex legal challenges. As governments and AI developers strive for innovation, a balanced and pragmatic approach that respects creators' rights and supports creative industries is imperative. For those seeking guidance on brand protection and intellectual property rights, BTO's Intellectual Property and Technology team stands ready to offer expert assistance.

Lynn Richmond, Partner & Accredited Specialist in Intellectual Property: lyr@bto.co.uk / 0131 222 2939

Laura Patriche, Trainee Solicitor: lpa@bto.co.uk / 0131 222 2939