- within Transport, Media, Telecoms, IT, Entertainment and Family and Matrimonial topic(s)
- with Senior Company Executives, HR and Inhouse Counsel
The Government has published its Report and Impact Assessment on the use of copyright works in the development of AI systems as required under section 136 of the Data (Use and Access) Act (the Act). The Act imported a requirement to produce a report dealing with the proposals put forward in the Government's consultation on Copyright and AI which closed in February 2025 to which no final and for which a report was also due. In today's report, the Government has concluded that now is not the right time to make concrete proposals in respect of almost every aspect of the copyright and AI issues considered.
This outcome, whilst frustrating, is perhaps not surprising, given that no clear consensus seems to have emerged from the consultation process between the creative communities and the Tech industry. The one area which did appear to have broad support was to address the transparency of use of copyright materials in the training and deployment of AI systems. However, the Government has not made any concrete proposals even on that point, seemingly due to the difference in nuance as to the preferred levels of transparency necessary. Instead, it is opting to "work with industry and experts to develop best practice on input transparency".
The one issue that is dealt with more positively is the potential repeal of the provision in the Copyright Designs and Patents Act 1988 which provides at section 9(3) that, "in the case of a literary, dramatic, musical or artistic work which is computer-generated, the author [and thus the copyright owner] shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken". This provision, which the consultation revealed was uncertain and not much relied upon, is also not present in any other jurisdiction's laws on copyright. Whilst some had suggested it might be used to provide copyright protection for works generated exclusively by an AI without human input, the Government's conclusion is that this could now be repealed. However, even this conclusion is less than definite, with the Report noting that "We propose to continue to monitor the use and impact of this protection...in the absence of evidence its ongoing value, we propose that it should be removed". The Government's report added the caveat that "copyright should continue to protect works created with AI assistance" ie human involvement may save copyright in AI generated works (where generated using AI as a tool not exclusively by AI). This proposal in itself therefore has unresolved tensions.
Last week, the House of Lords put out their own recommendations in anticipation of this report – see our blog post here – which backed a licensing-first model, rejected a commercial text and data mining (TDM) exception, and called for mandatory transparency on content used for training, protections for personality rights that could assist with deepfakes and "sovereign AI" (prioritising the development and adoption in the UK of models with transparency and copyright compliance meeting the Government's standards built in by design). There was little as concrete in the Government report today.
For a summary of the report's findings, see below.
HSF Kramer IP & Tech partner, Peter Dalton, commented:
"The Government Report on Copyright and Artificial Intelligence published today is notable for clearly articulating the challenges in the current copyright regime of balancing the interests of the creative industries and rights holders on one hand, and the Tech Sector AI developers and users on the other, whilst singularly failing to propose a way forward. This is perhaps not surprising, given that all prior efforts to find a working consensus between the creative industries and AI developers had failed, but has led to the Government abandoning its original proposal for a broad TDM ("text and data mining") exception for AI training, with an opt-out provision for copyright owners. However, one more definite point was the acknowledgement that the current s.9(3) CDPA, which was originally designed to deal with computer generated works over 35 years ago, is not clear and therefore was ineffective now the reality of AI generated materials has arrived, and so should be removed, rather than amended, if there is no evidence of ongoing value.
What also emerges is the support for some form of transparency over the use of copyright materials in the creation and training of AI systems. However, the Government has held off putting forward concrete proposals and instead will "wait and watch" whilst encouraging industry to find a practical way forwards.
The report does reflect the complexity of AI/copyright landscape, noting that even with copyright reform, the UK is unlikely to be the favoured jurisdiction for the initial creation of Foundation Models. The Government's stated focus is therefore on ensuring that those using AI systems downstream in the UK can do so appropriately, as well as noting the difficulties rights holders face where their material has been used for training outside of the jurisdiction, as reflected in the recent Getty Images case. The wait for a clearer framework on how copyright and AI work together will therefore continue."
Andrew Moir, HSF Kramer IP & Tech partner, added:
"While it's a good digest of the issues, the Government has largely decided to sit on the fence rather than risk upsetting either the creative industries or the AI tech sector. The "wait and see" approach, in a highly competitive international market where countries are competing for AI investment, also risks the UK being left behind as other countries take the first-mover advantage. Given AI models can be trained anywhere, it also won't prevent training of models on copyrighted materials, it will rather move the issue elsewhere."
SUMMARY OF THE REPORT'S FINDINGS
On the original proposal of a general TDM exception and opt-out option: No longer the preferred option
"In light of the strong views from the consultation, the gaps in evidence and the rapidly evolving AI sector and international context, a broad copyright exception with opt-out is no longer the government's preferred way forward. We propose to gather further evidence on how copyright laws are impacting the development and deployment of AI across the economy. We will consider and engage stakeholders on other potential policy approaches. We will also continue to monitor developments in technology, litigation, international approaches, and the licensing market." (para 27)
On transparency over content and data used to train AI: No specific proposals - the Government will continue watching what other jurisdictions are doing and work with industry to develop best practice
"We agree that greater transparency about how AI developers train their models, including the content and data they use, can help right holders assert their rights. We propose to continue monitoring the effects of transparency rules in other countries and consider our approach in the UK. Our approach must promote clarity and enforcement for right holders, without disproportionate effects on AI development or deployment in the UK. We propose to work with industry and experts to develop best practice on input transparency, to help right holders assert their rights. This will inform any future potential legislation." (para 30)
On labelling of human and AI-created content as such: No specific proposals - Continue watching what other jurisdictions are doing and work with industry to develop best practice
"We propose to work with industry to explore best practice on labelling AI-generated content. We propose to continue monitoring international developments and to work with international partners to support the development of common solutions". (para 33)
Development of technical tool and standards to help rights holders monitor AI's use of their work or licence it or prevent access to it: Effectively the same answer as above - monitor and work towards best practice with industry (para 36)
Licensing: No "intervention": the proposal is for monitoring and keeping market-led approaches under review. For AI developed outside the UK - the Government will continue monitoring global developments and judicial outcomes. It will also identify and assess "further leavers to support access to valuable datasets - including through a Creative Content Exchange" [see Section G of the Report]. (paras 40&41)
Enforcement: No specific plans - the Government will ensure enforcement in the UK is fit for purpose - consider regulatory oversight of transparency if related legislation is introduced
"We propose to continue working with partners, including law enforcement and the judiciary, to help ensure the UK enforcement framework remains fit for purpose. We propose further work to identify and address enforcement barriers and consider where action to mitigate these barriers may be required. We propose to consider the case for and approach to regulatory oversight of transparency or other measures, if legislation across these is introduced."
Computer-generated works: Proposal to remove s 9(3) CDPA but copyright should continue to protect works created with AI assistance
"Most people who responded to this consultation question considered that works created solely by AI should not be protected and supported [the removal of the protection under s 9(3) CDPA], while retaining protection for AI-assisted works. We agree that copyright should incentivise and protect human creativity. We propose to continue to monitor the use and impact of protection for wholly computer-generated works. However, in the absence of evidence of its ongoing value, we propose that this specific type of protection should be removed, while copyright should continue to protect works created with AI assistance." (para 47)
Digital replicas [Deepfakes]: Recognition that there are some legal protections in place today, but these do not cover all situations where a digital replica is made without consent. Proposal to:
"...explore a range of options for addressing these risks, while protecting the potential of this technology to support legitimate innovation. This exploration will include consideration of whether it would be beneficial to introduce a new digital replica or personality right." (para 49)
In parallel with the UK Government's lack of concrete proposals in this area, the EU is pressing ahead with detailed guidance on the obligations under the EU AI Act, in particular with regard to transparency. See our series of blog posts on the development of AI laws and regulation in the UK and EU here.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
