Hats off to actors’ union Equity. While Rishi Sunak pontificates and talks about talks with a view to taking on the threats and opportunities of artificial intelligence, Equity has already put together what it describes as a “ground-breaking AI toolkit”, which members can implement immediately.
The impetus comes from those members who have been inundating the union with requests for support amid concerns that the government’s “light touch” voluntary approach seems inappropriate, a stance taken only a couple of months ago that Prime Minister Sunak already appears to be attempting to U-turn.
The main concern is performance cloning. This is the process of creating a synthetic performance by recording, using or reproducing the performance, voice or likeness of an artist by machine learning systems and equivalent technology. The main threats identified so far are:
- Text-to-voice or image-to-voice translation or generation;
- Interactive digital humans or digital avatars capable of audio-visual interaction with users; and
- Manipulation of existing identities in audio-visual content such as deepfakes.
The concerns seem genuine, according to intellectual property expert Dr Mathilde Pavis, who helped the union to draw up the toolkit.
As she says, “the UK legal framework is not well designed to protect performers from unauthorised imitations of their work using AI technology. The Equity toolkit is a very good, but temporary, solution to protect performers until the UK Government reforms the law. We hope the government puts forward artist-centred and workable proposals for reform, going beyond voluntary codes of conduct. Having a clear legal framework on AI-generated performances makes for better business for all involved: performers, producers, content distributors, AI companies, and the consumer.”
As reported only last week, Equity is already struggling with challenges that members face around low pay and difficulties in claiming state support while training and “resting”. This is only going to get worse, as producers favour star-cast solo shows and cut back any acting or backstage role that they can, driven by a need to cut costs in an effort to survive the not-quite recession and cost of living crisis.
The last thing that any performer then needs is to find that AI has taken their job or, worse, to all intents and purposes stolen their identity and diverted much-needed income to a software provider.
Therefore, at breakneck speed, Equity yesterday released a model union contract, together with its AI toolkit and at the same time called on government to introduce an appropriate regulatory framework.
The intention of the toolkit is to set out ways in which AI can be used ethically at the same time as offering templates which attempt to ensure that artists can enforce their legal rights.
The toolkit comprises:
- A vision statement laying down ethical application of AI in the entertainment industry;
- A template AI contract relating to performance cloning work;
- Model AI clauses where performance is cloned without consent; and
- A template takedown notice for use where intellectual property rights are infringed.
This is an astonishing body of work to have put together in such a short time, when even governments and multinationals are, at best, posturing and concerning themselves with doomsday scenarios rather than the practicalities of dealing with AI in real life.
As Equity’s Industrial Official for New Media, Liam Budd, notes, “the government needs to step in with robust measures to properly regulate the growing use of AI.”
Therefore, in addition to taking the practical step of introducing the toolkit, Equity plans to approach the government identifying the urgent need to strengthen performers’ intellectual property rights and ensure that all entertainment engagers’ rights are protected and that they will have their images, voice and likenesses covered by GDP.