The Chronicles of AI

The cat is truly out the bag now when it comes to Artificial Intelligence or “AI”. AI is not a new concept by any means. The concept of computer AI has been around for over 70 years, but it is only in the last few years, and particularly with the release of ChatGPT by OpenAI in 2020, where AI’s progress has accelerated at such a pace that it has taken aback even the most ardent champions of the technology.

However, that does not mean that it has been smooth sailing. We are already seeing many complications over the use of existing content to create AI generated content and ownership issues of said generated content.

Recently, we have seen companies like Getty Images start legal action against Stability AI Inc for their alleged use of over 12 million Getty photos to train its Stable Diffusion AI image-generation system. This was followed by authors such as George RR Martin, John Grisham, Sarah Silverman, Mona Awad and Paul Tremblay being among a cohort of authors suing OpenAI over similar claims that their copyright was infringed in order to train ChatGPT without their permission.

OpenAI claim that the use of copyrighted materials by innovators in transformative ways does not violate copyright. However, this is a defence specific to the USA. In the UK it is not quite so clear. UK copyright law includes a ‘substantial part’ doctrine, which on some views allows transformative use. Moreover, UK copyright law allows a fair use defence, of which some commentators have seen transformative use as an element.

Thus, while UK law has not expressly recognised a transformative use defence, transformations may be allowed and could be considered fair use.

In a further move amongst concerned authors, in the UK the Publishers Association wrote an open letter to Rishi Sunak asking that the UK Government does not allow AI developers the free use of copyrighted songs and books for training. They want to see AI developers obtain necessary licenses in advance of any content being used. To that end the UKIPO is currently working with AI firms and rights holders to produce an agreement and guidance on copyright.

Whilst a balanced approach is likely necessary, as with most things, it remains to be seen if they can strike this balance and find a solution that both protects people in the creative industries and the development of the tech industry. However, one thing is for sure, we are all going to have to get used to AI and understand how it can help us but also what its limitations are and most importantly recognising when the content generated has come from copyrighted source material, something which is much easier said than done.

If you have any questions about copyright or your use of AI, don’t hesitate to reach out to us. Briffa is always on hand to help. We can be reached at info@briffa.com or on 0207 096 2779 for a free consultation.