Regulatory Background
The government seems to have finally taken more decisive action in light of the EU pressing ahead with its implementation of the EU AI Act[1], having previously refrained from promising any statutory provisions for the regulation of AI and specifically around the generative AI and protections of IP holders.
In February 2024, the former Conservative Government outlined its stance on AI regulation in its response[2] to the AI White Paper[3]. Their approach emphasised a flexible, sector-specific strategy to help strengthen the role of existing regulators. While acknowledging the future necessity for mandatory regulations on developers of advanced-general purpose AI models, they stopped short of proposing any plans for new legislation instead opting to await the development of case law and analysing the adequacy of existing regulatory framework.
Since taking office in July 2024, the new Labour Government has concentrated on fostering innovation whilst addressing the most significant risks posed by advanced AI technologies, marking a notable shift from the previous approach.
AI v Intellectual Property
In December 2024, the UK government unveiled its eagerly anticipated consultation on AI and copyright, introducing a comprehensive set of measures aimed at propelling the growth of both the AI sector and the creative industries.
The proposals seek to strike a balance by granting AI developers access to premium training materials, while simultaneously ensuring that content creators retain greater control over their work and receive fair compensation for its use.
The UK courts will also address the conflict between AI training models and intellectual property in the upcoming trial of Getty Images v Stability AI[4], scheduled for trial to determine liability on 9 June 2025. This case involves the use of the Getty image library to train the Stable Diffusion text-to-image deep learning model (a generative AI) and raises issues related to copyright, database rights and trademark infringement within the AI context.
The discourse only seems to be intensifying further with the latest launch of Sora by OpenAI which launched on 28 February to UK and European consumers. Sora is a video-generating AI which converts text prompts or images into up to 20-second-long videos in a range of ratios and resolutions. OpenAI has been less than cooperative in providing information on Sora’s training data with growing concern that unlicensed footage from video game playthroughs may be amongst the training data sets used.
How does the EU’s approach differ?
The EU has taken a less conservative approach to dealing with AI regulation having introduced the EU AI Act on 1 August 2024, with its provisions progressively applicable, starting 2025. The legislation takes the format of a risk-based approach classifying different AI systems into four levels: unacceptable risk (prohibited), high risk (strictly regulated), limited risk (subject to transparency obligations) and minimal risk (largely unregulated).
The first provision having taken effect on 2nd February 2025. This bought an absolute prohibition against certain AI systems which posed unacceptable risks to EU citizens’ safety and fundamental rights. Next will come the provisions on general-purpose AI models, which cover foundation models including those underlying generative AI systems and will be applicable from 2 August 2025.
What the Artificial Intelligence (Regulation) Bill [HL][5] seeks to introduce?
One of the main provisions of the Bill is the introduction of the ‘AI Authority’ whose functions are to address AI regulation within the UK. This will include, amongst others, ensuring that the relevant regulators take account of AI, coordinating a review of relevant legislation and assess its suitability to address the challenges and opportunities presented by AI, monitor and evaluate the overall regulatory framework’s effectiveness and support testbeds and sandbox initiatives to help AI innovators get new technologies to market.
The AI authority must adhere to certain core principles, many of which focus on the safety and security of individuals and data, ensuring transparency and accountability of these systems, and providing proper redress mechanisms.
Further to this it is clear that the Bill will attempt to address the current debate around the threat that AI models place on IP holders. The Bill (as currently drafted) states that “any person involved with training AI must supply the AI authority with all 3rd party data and intellectual property used in that training and assure the AI Authority that they use all such data and IP by informed consent; and that they comply with all applicable IP and copyright obligations”.
Critical Analysis – How can this affect IP rights holders and businesses?
This shows a substantial shift in the governments and legal professions’ proactiveness to AI regulation. This is possibly as a result of the growing discourse within the UK around IP infringement but also from market pressures in the EU and further afield to keep up with innovation and reform so as to keep the UK at the cutting edge of the AI market and boosting investor confidence by providing a clear and stable regulatory framework.
It is crucial to stay informed about upcoming legislative changes and emerging case law, not only for individuals and businesses within the creative industry and IP rights holders but also for companies involved in AI development and usage.
Clients will need to ensure they are aware of regulatory changes to keep on top of changing requirements, update contractual relationships to align with current AI provisions and avoid unnecessary disputes.
Additional Sources:
First rules of the Artificial Intelligence Act are now applicable | Shaping Europe’s digital future
Getty v Stability AI: A ‘tantalising glance’ of what’s to come for AI firms and creators
[1] Regulation – EU – 2024/1689 – EN – EUR-Lex
[2] A pro-innovation approach to AI regulation: government response – GOV.UK
[3] A pro-innovation approach to AI regulation – GOV.UK
[4] Getty Images and others -v- Stability AI – Courts and Tribunals Judiciary
[5] Artificial Intelligence (Regulation) Bill [HL] – Parliamentary Bills – UK Parliament