Skip to Main Content

Generative Artificial Intelligence and Copyright : Copyright and Ownership

A guide for University staff and students on the copyright implications of Using Generative AI tools.

Why is copyright relevant to developing and using AI tools?

Gen-AI tools are trained using materials (text, images, audio, video, code) which are protected by copyright. Use of third-party owned copyright works to train Gen-AI tools without permission from copyright owners could infringe copyright and there are several ongoing legal actions involving rightsholders suing technology companies for use of copyright works in training Gen-AI tools. There is also potential for a Gen-AI tool to generate something similar to a protected work by a real person and the likelihood of this giving rise to a claim for copyright infringement.  

There is uncertainty as to whether copyright protects AI generated works and if so, who owns the copyright with individual countries reaching different decisions. 

Does copyright exist in the output of Gen-AI tools?

It is not yet clear whether under UK law copyright protection will apply to images generated using Gen-AI tools.  For example, would a short text prompt to generate an image be sufficient to give rise to copyright protection? There is some debate over whether images generated in this way display sufficient originality to warrant protection.  For literary, dramatic, musical or artistic works copyright only applies if the work is the author's own intellectual creation.  

However, unlike many other countries the UK Copyright, Designs and Patents Act 1988 provides copyright protection for computer-generated works which do not have a human creator.  "In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken.” So, in UK law images generated using Gen-AI could give rise to copyright protection. 

What has still to be determined is whether ownership of copyright in an image generated using Gen-AI will lie with the company who created the tool or the individual user providing the instructions to create the specific output. In a landmark case a Chinese court held that an image generated using the open source AI software Stable Diffusion was protected by copyright finding that the image met the requirements of intellectual achievements and originality, should be recognised as a work and protected by copyright laws. The copyright owner was the individual who had input prompt words to generate the image. 

Who owns copyright in the output of Gen-AI tools?

The law is undecided as to whether copyright exists in the output of Gen-AI tools. UK law protects computer-generated works which could potentially include works generated using Gen-AI tools whilst US courts have found that AI generated works do not meet the threshold to constitute original works protected by copyright on the grounds that they lack human authorship.  

Where the law recognises copyright protection for such works then the copyright owner could be the creator of the Gen-AI tool who may have little connection with the creative input required to produce the output or the individual who inputs data (prompts) to produce a specific output. 

In the absence of legal certainty, it is especially important to check the terms and conditions of the Gen-AI tool that you use to determine where copyright ownership of output lies.