Why do we need transparency of copyrighted content in AI?

A critical debate striking at the heart of the UK’s creative industries has played out in Parliament as the UK positions itself as a global leader in artificial intelligence. On 10 June 2025, the House of Commons debated the government’s ongoing resistance to a key amendment proposed by the House of Lords to the Data (Use and Access) Bill. This amendment would have required AI developers to track and disclose the copyrighted material their models are trained on, ensuring greater transparency in AI development.

However, on 12 June 2025, after the Bill had been passed back to the House of Commons from the House of Lords, the government refused the amendment and the Bill has been passed and will become law once Royal Assent is given.

The debate marks a pivotal moment for freelance artists, software developers, game designers, and other creatives whose livelihoods depend on the protection of their intellectual property.

What’s at stake for creators?

At the heart of the Lords’ amendment was a call for transparency, a principle that resonates deeply within the creative sector. AI models are often trained on large datasets scraped from the internet, frequently without the consent, credit, or compensation of the original creators. According to Creative UK, the UK’s creative industries contributed £123.03bn in gross value added during 2024. For a sector of this magnitude, unregulated data usage is not only ethically questionable but also economically risky.

The government’s current stance is that they are going to allow a consultation on the issues conclude and that may lead to a separate AI Bill. The way that it progressed after the latest rejection by the House of Lords is that there is to be a report published by the government on its copyright and AI proposals within nine months of Royal Assent and there is to be an interim report published within six months.

This all comes at the same time as the Getty Images v Stability AI court case in the UK which will consider the allegations of copyrighted images being used in the training of Stability’s AI mode and the reproductions of substantial parts of that copyrighted content in the outputs of the model. The outcome of this landmark case is likely to be known before the outcome of the government’s consultation on the training of AI models on copyrighted content.

What is the proposal from the Copyright Licensing Agency?

An interesting development has come from the Copyright Licensing Agency (CLA), which plans to launch its own “Generative AI Training Licence” later in 2025. The CLA states that this will “ensure” remuneration for publishers and authors, while offering AI developers the legal certainty to use a wide range of content to train language models.

However, details remain unclear. It’s not yet known how the licence would be implemented, especially in the absence of a mechanism for creators to opt out of AI training models or how AI developers would identify assets which would be subject to a licence.

Will these proposals bring balance?

Baroness Kidron (a former film director and crossbench member of the House of Lords) has described the government’s approach as bordering on “state-sanctioned theft.” She had warned that ministers would be “knowingly throwing UK designers, artists, authors, musicians, media, and nascent AI companies under the bus.”

These concerns are not exaggerated. The UK risks undermining the very industries it seeks to empower through innovation if it does not implement strong, enforceable protections.

The Lords’ proposed amendment not only demanded disclosure from AI companies about what copyrighted content is used in training, but also called for a statutory review within 15 months of the Bill’s enactment to assess its impact on the creative economy. This would act as a crucial safeguard for creators’ rights amid the rapid advance of AI technology.

Can innovation and integrity coexist?

The government has warned that stricter copyright regulations could deter AI investment and stifle innovation. But this presents a false choice. Innovation and intellectual property rights can coexist.

It is entirely feasible to create a legal framework that protects creators’ rights while promoting ethical and sustainable AI development.

By ensuring fairness and transparency in AI training, the UK can attract businesses committed to responsible innovation where creators are not seen as obstacles, but as essential collaborators.

What does the Data (Use and Access) Bill mean for our clients?

At Eaton Smith, we represent a wide range of clients across the creative industries - from game developers and freelance artists to musicians and software programmers. We understand the critical value of creative work and the need to protect it in an increasingly automated world, where intellectual property is constantly at risk of unlicensed use online.

AI and its use in game development is a common theme that arises in discussions with clients and events at which we attend. Recurring topics are ownership of AI-generated content and the conversion of text-based prompts to artwork, the current position of the UK’s Copyright, Designs and Patents Act 1988 with regard to the author of the output and also the Terms of Service of the particular AI service that is being used. There is the risk that outputs that are based on existing works upon which a large language model may have been trained and which might as a result form a part of the output, may infringe the intellectual property rights of any third party intellectual property rights holder, such as is being tested in the Getty Images v Stability AI case referred to above.

We believe that any future AI legislation must put creators at its core. Transparency, consent and fair compensation are not optional; they are fundamental to a thriving creative economy.

While AI offers significant economic potential, that value should not come at the expense of established creative sectors.

Our clients and the wider creative community will need to monitor future AI regulation closely to ensure work is protected and fairly compensated.

We’ll be watching closely as this important issue continues its circuitous journey through Parliament, the courts and the consultation period.

Disclaimer: Nothing in this piece is legal advice and the issues covered are non-exhaustive. It is intended to provide information of general interest about current legal issues. As all circumstances are different, you should take specific legal and any other relevant professional advice before acting in reliance on any of the information provided.