The UK government is about to launch another consultation on copyright and generative AI. As part of this consultation, ministers are expected to propose the introduction of a new personality right into UK law. In this TW News Briefing we explain what that means and why it matters for performers.

---

GENERATIVE AI RECAP
As outlined in this explainer, the term generative AI refers to AI models that are able to generate original content, including text, image, audio, video and music.

These models are trained by being exposed to existing content. So AI companies copy large quantities of existing content onto their servers to form a training dataset.

That training data is often copyright protected content owned by third parties. In order to copy that content onto their servers, the AI companies should get permission from – and therefore negotiate a licensing deal with – all the relevant copyright owners.

But quite often that doesn’t happen. Even though the creative industries are adamant that permission is required.

Many AI companies argue that they don’t need permission because they can rely on data mining exceptions that exist in some copyright systems – or the concept of fair use under US copyright law.

The creative industries disagree and this dispute is at the heart of numerous lawsuits that have been filed by copyright owners against AI companies.

UK copyright law doesn’t currently have a data mining exception. However, it is thought that the government is considering introducing one, albeit with the option for copyright owners to opt out, and that will be one of the proposals set out in this week’s consultation document.

This is a controversial proposal that we discussed in this previous briefing.

The government is obviously aware that that proposal is controversial and will result in a strong backlash from the creative industries. Reports suggest that ministers are hoping that also proposing the introduction of a new personality right might placate the creative industries to some extent.

---

LEGAL QUESTIONS POSED BY DIGITAL REPLICAS
Some generative AI models are able to generate content that imitates the voice or likeness of a specific individual. These outputs are sometimes referred to as digital replicas. These models pose the question: can individuals stop the unauthorised imitation of their voice or likeness by generative AI?

This control is partly needed so that performers and celebrities can control and commercialise the use of their voice and likeness in this context, if they so wish.

However, there is also a wider need for this control. Basically so that anyone and everyone can stop AI being used to make it appear like they said something they didn’t say or they did something they didn’t do.

In order to train a generative AI model to imitate an individual’s voice or likeness, the model will need to be exposed to a training dataset that includes audio or video featuring that individual.

That content will usually be protected by copyright. And as far as the creative industries are concerned, this content can only be used with the explicit permission of the copyright owner. However, this doesn’t necessarily provide protection for the individual’s voice and likeness.

First, as stated above, many AI companies argue that they don’t need permission from the copyright owner because of copyright exceptions or fair use.

And second, even if they do need permission from the copyright owner, the individual may not own the copyright in the audio or video that features their voice or likeness. For example, the copyright may be owned by a film studio, TV producer, audiobook publisher or record label.

Therefore it is generally agreed that another right is required to protect an individual’s voice and likeness in the context of AI.

---

PERSONALITY RIGHTS
Many countries provide what are variously known as personality, publicity and image rights.

Quite how these work differs around the world, but they generally allow an individual to control the use of their likeness and, usually, their voice.

It means if a third party wants to exploit an individual’s likeness or voice they need to get permission, otherwise they can be sued for infringing the individual’s personality right.

No such right currently exists in the UK. Instead, where a third party exploits an individual’s likeness or voice without permission, the individual would have to sue for so called passing off.

This is a law that stops companies misrepresenting to their customers that they have a relationship with or endorsement from another party, when in fact they do not.

Although passing off has been used in disputes in the UK where publicity rights would be relied upon in other countries, it doesn’t really provide the same level of control or protection.

With is why, as part of the debate around generative AI, many creator and performer groups in the UK have been calling for a new personality right to be introduced.

---

HOW MIGHT A NEW PERSONALITY RIGHT WORK?
It remains to be seen how the personality right being proposed by the government might work. However, the UK could look to the US for guidance.

Currently personality rights in the US exist at a state level rather than a federal level, which means that rules differ around the country.

However, a new US-wide digital replica right has been proposed in US Congress and the US Copyright Office recently published a report on how such a right might work. Those proposals and that report could be influential as personality rights are discussed in the UK government’s new consultation.

For performers who work with studios, producers, publishers or labels, one key concern will be to what extent those business partners can exploit those performers’ personality rights under both existing and future deals.

Performers who want to commercialise their voice or likeness in the context of AI will want to work with business partners to help make that happen.

However, there are also concerns that those business partners might try to negotiate deals that given them long-term wide-ranging control over the new right, which may not be in the best interests of the performer.

The proposals in US Congress include some protections in this domain, by restricting what business partners can do when they enter into deals relating to an individual’s voice and likeness.