The Blue Highway Beacon: Vol. II, NO. 7

How Can We Reward the Talent When Their Creations Are Used in AI-Generated Works?

By Amanda Harcourt, Honorary Professor of Practice, IBIL, Faculty of Laws, University College London, and Head of International Copyright, Contracts and Rights Administration for Blue Highway

The Impact of AI on Creative Industries

Shall we call out moral panic or summon the pearl-clutchers?

In music, film, writing, and art, Artificial Intelligence (AI) and its cousin Generative AI invite conflict and concern. Authors and performers rightly see that the powers in Silicon Valley, having already decimated their incomes over the last few decades, seem intent on a repeat performance. From the talent’s perspective, regarding AI and its clever coding, Silicon Valley titans proceed based on the old principle of forgiveness rather than permission. The same old swindle continues, and the talent has a legitimate point.

The Challenges of AI and Copyright

The creative industries face several problems today. One is whether the training of AI infringes copyright by copying. Google seems not to think so, but many disagree. Numerous voices assert that the U.S. legal principle of fair use applies when training AI. This position has been a topic in U.S. academic discussions and courtrooms for some time. Is this principle what led to the revelation that Meta employees, with apparent approval from Czar Zuckerberg, downloaded approximately 82 Terabytes of pirated books from shadow libraries to train AI systems.

To complicate matters, Meta has announced that it will be using all public user data, including texts, photos, and comments, to train its AI from June 2026. This practice has already been in play at Facebook and Instagram since 2007!

Legal Perspectives: Fair Use and Fair Dealing

What constitutes fair use? You must decide. The fair dealing regime in the U.K. is clearer but still at risk of abuse. U.K. law provides a closed list of actions that do not require a license. Under E.U. law, text and data mining (TDM) exceptions to copyright are limited (as outlined in Article 3 of the 2019 DSM Directive).

TDM is permitted when it involves reproductions by research organizations and cultural institutions for scientific research purposes, as they have lawful access. It is not acceptable when Meta is building a massive AI tool in the E.U., then.

Next, we confront the issue of whether authors and performers should have the option to opt-in or opt-out of having their works used to train AI. The latter is more complex due to the "that ship has sailed" concern. Furthermore, should the decision to opt fall solely to creators or also to their commercial partners?

Payment and Revenue Distribution

Another pressing concern is whether the money generated from AI models trickles down from corporations to creators. The U.K. government has initiated an AI consultation through the U.K. Intellectual Property Office. However, this effort seemed to many like a fait accompli due to the Government’s intention to establish a strong British AI industry.

Creative industries expressed fury at hints that the U.K. Government would introduce new copyright and database exceptions allowing TDM for any purpose. Fortunately, the idea of an opt-out has been proposed, but as a respected British copyright academic noted, “[w]hether it makes sense or not, is a different thing.”

The E.U.'s approach, though not perfect, has proven more flexible. For instance, French and German collective management organizations for music authors have issued proceedings against AI infringers. They have also published intentions to opt-out and developed a licensing framework. The Irish music CMO, IMRO, has taken a further step, allowing members to opt-in via a meticulous process.

Sadly, the influential U.K. music CMO has not followed suit. PRS for Music appears distracted, dealing with a trust and confidence crisis.

Identifying AI-Generated Works

How do we differentiate between AI-generated and human-generated works?

A recent U.S. Copyright Office study from January found many submissions aligned with U.K. and E.U. legal principles, advocating that the current legal regime is adequate. It emphasized that free and creative human choices remain crucial for copyrightability. An AI-generated work based on an existing copyright work can, theoretically, be protected only to the extent of the copyrighted material within it.

This aligns with the author's stance taken in Ontario last autumn during Canada’s International Institute of Communications Conference.

The Role of Consent and Reward

From a practical perspective, the question remains: How can we ensure that authors and performers, who have rights, are part of the consent-and-reward equation concerning AI?

The Berne Convention Article 5 (2) states that “The enjoyment and exercise of copyright is not subject to formalities.” Copyright exists once an original work is expressed in material form. The initial copyright owner is the creator or author. Although ownership might change, authorship does not. The owner must be consulted and rewarded when their work is used. This presents a significant issue. In the words of George Merrill and Shannon Rubicam for Whitney Houston, “How Will I Know?”

Identifiers for Copyright Work

A global standard identifier exists for each copyright category. The ISO Office allocates identifiers globally. For music, ISWC is for songs and ISRC for recordings. Books carry an ISBN on their spines. An audiovisual work should feature an ISAN, but Hollywood, the BBC, and others choose to use their own, non-ISO labels.

Encouragingly, in May of last year, 29 countries adopted the ISO's ISCC, an open-source technology for identifying and managing digital content. This includes a unique, decentralized identifier for various media types. This is great news if—and only if—everyone uses it.

Revisiting Registration Requirements

Despite these advancements, the absence of a central registry for authorship and ownership of global works raises questions. Should we reconsider the “no registration required” clause in Berne?

An important discussion will occur in the U.K. on Wednesday, May 14th, 2025, at 1730 British Summer Time. The Institute of Brand and Innovation Law, part of the prestigious Faculty of Laws at University College London, will host this public event.

Access is available both in-person in the UK and online.

Join us at the event!

I will chair the event, featuring stellar panellists with diverse perspectives and backgrounds.

Key Panellists

For the U.S., UCL is honored to welcome Jalyce Mangum, Attorney-Advisor at the Office of the General Counsel, U.S. Copyright Office.

Krishna Sood, Assistant General Counsel at Microsoft, will provide insights into these registration challenges.

The role of collective management of copyright works in this context will be presented by Adriana Moscoso del Prado Hernández, General Manager, GESAC for European Authors’ Societies.

Finally, we will have Professor Martin Senftleben, Director of the Institute for Information Law (IViR), University of Amsterdam, offering an academic viewpoint.

Fees for this event are remarkably affordable, ranging from £8 to £35. You can book tickets through this link.

© Amanda Harcourt 2025

Honorary Professor of Practice, IBIL, Faculty of Laws, UCL

https://harcourt.global/ & at Blue Highway Advisory