Trusted legal advice since 1733
Blandy & Blandy Solicitors

Season's Greetings and Christmas Closure Information Read more >

Insights // 05 January 2023

Artificial Intelligence, Copyright and Your Selfies

Partner Debbie Brett, head of our Corporate & Commercial team, looks at some of the emerging issues surrounding artificial intelligence.

AI has been creating a buzz recently. When scrolling through social media you may have noticed your connections’ selfies and profile pictures being reimagined in different art forms, from cartoons to renaissance portraits. This has opened up many debates, with some claiming that the AI generated selfies are discriminating against protected characteristics (e.g. creating hypersexualised images of women, compared to the images generated of men and generating images of racists stereotypes). There have also been concerns about privacy and how the selfies that users are uploading are being used (NB: always read the privacy policy!).

Another issue that has sparked debate is whether AI is ‘stealing’ the copyright in artists’ work by reproducing images in the same style, or using elements of their work, without their consent. This is linked to the recent ruling in the court of Luxemburg in which American based photographer, Jingna Zhang, pursued a claim against painter, Jeff Dieschburg, in which it was alleged that Dieschburg had copied one of Zhang’s images. You can see a side-by-side comparison of the photograph and painting, that Zhang posted on her Instagram feed (@zemotion) on 1 June.

Zhang was not successful in her claim. The Luxembourg District Court ruled that the photo did not infringe on copyright due to lack of originality in the photo, as the model’s pose was not ‘unique’. Under English Law, for works to benefit from the protection of copyright (as set out in the Copyright, Designs and Patents Act 1988 (CPDA)), the works must be:

a) Within a category of protected work, being: a recording of literary, dramatic or musical works; sound recordings, films or broadcasts or a typographical arrangement of published editions; and

b) Original, meaning that ‘the author must have created the work through their own skill, judgment and individual effort and that it is not copied from other works’.

But what is the position with AI under English Law? Some users have claimed that their generated selfies even show watermarks and artists’ signatures in the images the apps have created. Most AI apps that generate these types of images are programmed to create new images based on altering existing images (usually those collected from the internet). The existing images that the AI learns from is referred to as ‘training data’, which teaches the AI to do its job – in this case, to generate artistic selfies.

The CPDA explicitly protects computer-generated works which do not have a human creator, so in theory the artwork created by AI can benefit from copyright protection. But who would be liable for infringing the copyright in the existing images used as training data?

This leads to the ‘text and data mining exception’ under section 29A, CPDA which ‘provides that the making of a copy of a work by a person who has lawful access to the work does not infringe copyright in the work provided that (i) the copy is made to carry out a computational analysis of anything recorded in the work for the sole purpose of research for a non-commercial purpose, and (ii) the copy is accompanied by a sufficient acknowledgement (unless this would be impossible for reasons of practicality or otherwise).’

In conclusion, the developers of AI would need to obtain a licence from the authors of any works that are protected by copyright before using such works as training data, unless the s. 29A exemption applies, otherwise they run a risk of infringing a third party’s copyright.

There is no doubt that the development of AI has, and will continue to bring, challenges for law and policy makers to keep up with AI and how it is used and commercially exploited. As this is an uncertain, and often contentious area, the use of AI, the ownership of any training data and output created from AI should always be clearly set out in a contract before businesses enter into commercial endeavours involving AI.

For further information or legal advice, please contact law@blandy.co.uk or call 0118 951 6800. 

This article is intended for the use of clients and other interested parties. The information contained in it is believed to be correct at the date of publication, but it is necessarily of a brief and general nature and should not be relied upon as a substitute for specific professional advice.

Debbie Brett

Debbie Brett

Partner, Corporate & Commercial Law

Read Bio