16 February 2024

In March 2023, following a review on pro-innovation regulation for digital technologies by Sir Patrick Vallance, the UK government announced plans to seek to clarify the relationship between intellectual property and generative AI. To do so, the government sought to work with users and rights holders on a code of practice on copyright and AI, which it hoped would strike a balance between AI developers’ requirement to access data to train their AI models, and content creators who have the right to control access to their works. The first working group meetings were held in June 2023.

Recent media reports, however, indicate that those group meetings have failed to yield a workable solution. As such, it has been reported that the responsibility for creating a code of practice has been returned to the Department for Science Innovation and Technology.

A blow to the creative industries

The decision is likely to come as a blow to the creative industries, which have long expressed concerns about their works being used to train AI models without their consent.

In a recent report by the House of Lords Communications and Digital Committee, the Copyright Licensing Agency argued that current practices by technology companies “severely undermine not only the economic value of the creative industries but the UK’s internationally respected ‘gold-standard’ copyright framework”. Similarly, the CEO of the Publishers’ Association argued that large language models are “infringing copyright on an absolutely massive scale”.

At home and abroad, these tensions have led to court action. In the US, authors including John Grisham, Jodi Picoult and George R.R. Martin are amongst 17 authors currently suing Open AI for “systematic theft on a mass scale”, which infringement has been described in the court action as “flagrant” and “harmful”.

In the UK, a similar litigation is ongoing between Getty Images and Stability AI, with Getty initiating legal proceedings and alleging that Stability AI has infringed its intellectual property rights by unlawfully copying and processing 12 million photographs and associated captions and metadata without permission (see our article on that here).

In response to the action, Stability AI asked the High Cout in England to strike out two aspects of Getty’s claims, so that they would not be considered at a full trial: the training and development claim (which is highly relevant to rights holders and would at least in theory shed light on the extent to which this constitutes infringement); and a claim of secondary infringement in relation to the availability of Stable Diffusion (the system at the heart of the dispute) in the UK. In December 2023, however, the High Court confirmed that the issues are both sufficiently contested and unclear that they should not be struck out. 

The dispute looks set to go to trial and should provide welcome clarity from the judiciary on the extent to which AI systems infringe copyright by using copyrighted materials to train data.

The legal position

It is worth noting that the Copyright, Designs and Patents Act 1988 (“the Act) provides that in the case of a literary, dramatic, musical or artistic work which is computer-generated, the author is the person by whom the arrangements necessary for the creation of the work are undertaken.

There is a case to be made that the person “by whom the arrangements necessary” are undertaken is the developer of the AI software, but also the person inputting the data to yield a particular outcome. From a practical perspective, a rights holder looking to sue for copyright infringement would be best advised to seek recourse from both.

Nevertheless, it is also worth being aware that even if the question of ownership can be put beyond doubt, the question of originality still requires to be answered. In the UK, a work will generally only be classed as “original” if it results from the author’s free and creative choices and exhibits his or her personal touch or personality. There is no definition of “original”, which is fact dependent and a question of degree but broadly the legal position is that the work is the author’s own intellectual creation.

Looking ahead 

The future of the relationship between AI-generated works and rights holders is difficult to predict and there is little doubt that both parties would benefit from clarification in relation to copyright protection for AI-generated works.

Achieving a scenario that works for both will require a delicate balance of competing interests and clearly requires more consideration than it has thus far been given.

The code of conduct discussions have stalled for now but presumably will be taken forward by the Department for Science Innovation and Technology and updates are awaited with interest.

Lauren McFarlane, Associate: lmf@bto.co.uk / 0131 222 2939