On 17 December 2024 the UK Government announced its public consultation on 'Copyright and Artificial Intelligence' (Copyright and AI Consultation) in which it seeks views on its proposal to make significant changes to the UK's existing copyright regime by introducing an EU style broad text and data mining (TDM) exception that would apply to training AI models.  The UK Government has said that its proposal aims to:

• "enhance right holders’ control of their material and their ability to be remunerated for its use;
• support wide access to high-quality material to drive development of leading AI models in the UK; and 
• secure greater transparency from AI developers to build trust with creators, creative industries, and consumers."

Notably, the Copyright and AI Consultation also goes beyond copyright and databases rights to touch on personality, privacy and performers rights – something we touched on in our webinar on "The Impact of AI on the Media & Entertainment Industry" with Lord Clement-Jones – former Chair of the House of Lords Artificial Intelligence Select Committee and currently Co-Chair of the All-Party Parliamentary Group on Artificial Intelligence (see here) . It also flags the importance of AI in education.

Current UK TDM Exception

Section 29A of the Copyright Designs and Patents Act 1988 (CDPA) provides, following amendment in 2014 (well before current debate on AI), for a copyright exception which allows TDM for non-commercial purposes only and in circumstances where: (a) the user has lawful access to the copyright work (e.g. by virtue of a licence or permission in the terms and conditions); (b) there is sufficient acknowledgment; and (c) it is otherwise dealt with in accordance with the rules contained in this provision (such as the requirement for the copy of the copyright work to have not been transferred to another person). This means that copyright works protected under the CDPA cannot be used for commercial TDM purposes (such as to train AI models) without the authorisation (such as a licence or a consent) of the rights holder unless one of the other exceptions under the CDPA applies.

Previous proposals to reform the UK TDM exception

The latest proposal follows a previous (now abandoned) proposal published by the UK IPO in June 2022 in response to its consultation on "evidence and views on a range of options on how AI should be dealt with in the patent and copyright systems". In the response, the UK IPO indicated that it planned to broaden the TDM exception under the CDPA to allow TDM for any purpose (including commercial one) without granting rights holders any ability to "opt out". In February 2023, the UK Government u-turned on its plans after the proposals drew widespread criticism from rights holder groups, particularly within the creative industries, over loss of revenues under TDM licences given that some rights holders will have existing established TDM licensing arrangements in place.

EU approach and the UK Government's Proposal

In the Copyright and AI Consultation, the UK Government now proposes (a middle ground between the existing narrow TDM exception and the 2023 UK IPO Proposal) to introduce an exception expressly recognised to be similar to that which applies in the EU, as provided by Article 4 of the Digital Single Market Copyright Directive (Directive (EU) 2019/790) (TDM Directive), which allows reproductions and extractions of lawfully accessible copyright works and database rights for the purpose of TDM generally, including for commercial purposes. As with the exception under Article 4, DSM Directive the new UK TDM exception currently proposed would apply to all content unless the rights holder has pro-actively taken steps to “opt out” of the exception by expressly reserving their rights.  In other words, as under the TDM Directive (and now also confirmed in the context of generative AI in the EU AI Act), in the absence of an opt out, in content (including that which is protected as a copyright work or as database rights) which is available on the internet could be freely reproduced and extracted including for commercial TDM purposes, such as to train a foundation model. This would impact the revenue that could be generated from copyright licensing and obviously presents a risk if rights holders are not aware of the exception and fail to expressly “opt out”.  Especially given that, once the data is extracted from a website, the duration of the TDM exception is unlimited. 

As with Article 4, DSM Directive in the EU, an effective opt out would therefore be crucial if rights holders wish to reserve their copyright and database rights. However, the UK Government also acknowledges in the Copyright and AI Consultation that, to date, the "opt out" method required by Article 4, DSM Directive (which requires the rights holder to reserve their rights "in appropriate manner" such as by machine readable means) has left  "uncertainty about how it works in practice and some aspects are still being developed" and expressed concerns that the practical application of the "opt out" has not been consistent. However, no solution for tackling this has been proposed by the UK Government at this stage and it has instead welcomed suggestions from respondents to the consultation as to how these issues might be resolved. The UK Government considers it likely that collective rights management organisations will play a key role in developing licensing solutions in this area (as some already are).   

The UK Government also confirmed that the proposed wider TDM exception would be underpinned by greater transparency about the sources of training material and content generated by their models and that it will not introduce a new TDM exception without enhanced transparency measures (which may require legislation to provide legal certainty). These could include requirements for AI firms and others conducting TDM to: (i) disclose the use of specific works, datasets and web crawlers; and (ii) keep records, to provide certain information on request, or to evidence compliance with rights reservations.  The Copyright and AI Consultation seeks views on these obligations and also the EU's approach to transparency under the EU AI Act. See our previous article Training AI models: Content, copyright and the EU and UK TDM exceptions for a summary of those obligations.

Deepfakes 

In addition to TDM, the Copyright and AI Consultation also seeks views on whether the UK's existing patchwork of intellectual property and related rights (which includes the common law tort of passing off, performers rights under Chapter 2, CDPA and copyright (such as in photographs and recordings) provides adequate protection to performers or whether this needs to be updated to better protect against the unauthorised use of their image, video and/or audio in the inputs and outputs of AI tools to create deepfakes (aka "digital replicas"). In particular, the UK government refers to the July 2024 US Copyright Office report on digital replicas which recommended the introduction of Federal Digital Replica Law and the California Assembly Bills (2602 and 1836) which seek to provide specific protection to performers in the context of AI-generated digital replicas.

Computer Generated Works

The Copyright and AI Consultation also seeks comments on the effect of the specific protection afforded to computer generated works by s. 9(3), CDPA which provides that where 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 (CGW Protection). In particular, it focusses on two areas of criticism of the CGW Protection, that: (i)  it is incompatible with the modern legal test for originality which requires that a copyright work be an "author's own intellectual creation" being an expression of their creative choices and reflecting "personal touch" (as per the CJEU ruling in Infopaq C-5/08 and confirmed by the UK Court of Appeal in THJ Systems Ltd v Sheridan [2023]) which focusses on human qualities that a machine-authored work could not satisfy; and (ii) CGW (as a distinct category of protection) is either not needed, has no effect or has a detrimental (moral and/or economic) effect.  The UK Government has indicated that its preference is to remove this specific protection in the event that responses to the Copyright and AI Consultation "reveal insufficient evidence of positive effects from CGWS protection".

What can we expect next?

The Copyright and AI Consultation is now open to public consultation and will close on 25 February 2025.  Once closed, the UK Government will consider the responses and publish its response which is likely to outline its plans and next steps. Given the scope and nature of the consultation it is likely to be only the state of further engagement ahead of any draft legislation.

Like to read more on this topic?

We have previously written about the UK TDM and EU TDM exceptions in the context of AI  here: Training AI models: Content, copyright and the EU and UK TDM exceptions, Ally Clark, Duncan Calow