Some significant changes relating to issues around AI are on the horizon.
In today’s global market, keeping ahead of the different AI regulations across multiple territories is tough.
Ethical concerns are being raised over how it is being used in the creative industries following stories such as Chinese actors being replaced by AI doubles using TikTok owner ByteDance’s Seedance 2.0.
Responsible use of the new tech is paramount, not just for public service media, but for producers who want consumers to trust their content.
More regulation and guidance are also due to be discussed over the next few weeks in the UK, most immediately the issue of copyright in a talk between Adolescence writer and Writers’ Guild of Great Britain (WGGB) president Jack Thorne and Baroness Beeban Kidron on Sunday at London’s South Bank Centre.
Further announcements are due to be made too by bodies such as AIMICI, which is working on the issue of how consumers might be informed about how AI has been used in the making of programmes they are watching.
To help you through this shapeshifting topic, this week we’ve spoken to some of those at the forefront of the evolving conversation and lobbying in the UK and Europe around AI regulation and guidance, including AIMICI, the European Broadcasting Union, union Equity and WGGB.
Read on for snapshots about:
The latest thinking on monitoring and disclosure of AI use to the industry and viewers
AI protections in the PACT/Equity agreement and regulation around ‘personality rights’
How the EU AI act and new SAG-AFTRA deal might affect you
Copyright and government lobbying
Monitoring and disclosing AI
At the Edinburgh TV Festival Particle 6 chief executive and AI character Tilly Norwood creator Eline van der Velden was refreshingly frank about the use of AI, joking that, “saying you’re working with AI is like saying you have an STD” as “no one’s wanting to admit that”.
The fear of reputational damage by using technology that could replace humans and put humans out of work is real.
But AI is here to stay in various shapes and forms (van der Velden said she’s signed “hundreds of NDAs” with organisations wanting to use her AI expertise) so what can you do to use it responsibly?
AIMICI - which bills itself as the “leading responsible AI governance organisation for the screen industry” is at the forefront of the latest thinking and training.
Its managing director Kathryn Webb said China’s use of digital doubles to replace actors might seem a long way away but she stressed the “global element” of AI.
“Conversations in the UK and even in the US sometimes do often go, ‘Oh China’s it’s own thing…they’ve got their own way of approaching things’ but the reality is it’s an interconnected world, and often…the highest quality models available are coming from China, and they may continue to come from China.”
She pointed out China does have some regulations too that balances the technical aspect of all that is possible through AI, but said that, “it’s not surprising to me that where the technology has started to come in” is in low-budget, vertical video micro-formats because “the focus is very much on cost saving, not so much on creative excellence.”
Webb says, “people will continue to use AI and our hope and our hypothesis in terms of what we do is we focus on two key things to try and help make that be done in a more responsible way.”
The first thing is training in using AI ethically, which AIMICI provides, plus advice on governance and strategy.
“But essentially you can’t really have these conversations until everybody’s on the same page about what these different technologies can do, what they are, where the money goes and also what kind of checks and balances can be put in place to help prevent any misuse or irresponsible use.”
In addition, AIMICI is workshopping how the screen industry can report their AI use and signal it to viewers on-screen.
Details about that are due to emerge very soon.
Webb explains: “We’re encouraging a kind of unified, industry standardised type of disclosure…to put onto production”, similar to the BAFTA Albert sustainable production monitoring and reporting system.
“The misconception is that you can simply share a minimum amount of information about what the AI is and that kind of does the job. I think that there is a lack of good process and industry standardised process within productions about documenting and communicating AI use.”
She adds that AIMICI has also created “this kind of tiered system of documents, so that you can still have it as an internal reference; you have something that can be shared with distributors and others, if needed.”
Takeaways: keep up to date with global AI regulation as industry is interconnected; if you’re not already, start implementing the monitoring and recording of your AI use
PACT/Equity agreement and personality rights
Something to be aware of is that union Equity is close to securing a brand new clause around AI protections and uses in the forthcoming Equity/PACT Agreement.
Representatives from Equity were in California earlier this year observing and supporting US union SAG-AFTRA in their negotiations where they secured some protections on AI.
Expect some of that to feed into the new Equity PACT Agreement, which should be going out to members this autumn and will be the first time AI protections for performers will be written into a collective agreement, along with usual things like holidays, pay etc.
Also, more widely, Equity is calling for greater protections to be brought in, not just for actors, but for everybody.
The union thinks the government is due to open a consultation this autumn on ‘personality rights’, which it has been specifically asking ministers to do.
Equity has done a comprehensive policy briefing on the issue, that lays out how to protect both performers in terms of their likeness, voice and movement, but also protect anybody’s likeness or voice from being scraped for digital avatars.
“We believe people should have the rights over themselves, so they can’t be turned into deep fakes online etc” explained an Equity spokesperson.
From an industrial point of view, this is actors’ work: their bodies, voices and motion are the tools they use to create their performance and sell their work.
Another thing to note, if you’re considering turning some of your IP into a video game, the video games industry does not have a collective agreement with Equity, despite receiving tax relief from government.
The union has heard of “horror stories” from actors doing motion capture work in studios and, in order to protect their personality rights, believes tax credit arrangements that exist should be linked to minimum terms and conditions for workers in that industry as well, so watch that space.
Also, increasingly there are companies, such as Synthesia, where you can buy AI avatars for your content..
Do your due diligence which outfit you choose if you are buying a digital double as Equity has heard of some actors who have sold their likeness via motion capture to disreputable AI operators and found their avatars “end up promoting content that they would never in a million years have done in real life.”
So check the avatar you’re buying is from a company that creates and distributes them responsibly.
Takeaways: AI protections are due to appear in the forthcoming PACT/Equity agreement; a government consultation on ‘personality rights’ is expected
Overseas AI regulation and deals
The key agreements to look at are the AI protections in the new SAG-AFTRA deal and the EU AI Act.
In the US, SAG-AFTRA members have approved a new four-year contract that means AI performers are only allowed to be used on a production if they bring “significant additional value” compared with a real actor or their digital double.
Webb thinks that, “a lot of people have been kind of caught off-guard” by the EU AI Act “because it’s been talked about for so long” so “we’ve been working to translate that down into, as a producer, as an actor or as someone working on a production, what does that mean and what does it mean now for your day-to-day?”
The act came into force in 2024 but it is being implemented in phases. An important one began in August that covers transparency. According to the EU, the AI Act “introduces specific disclosure obligations to ensure that humans are informed when necessary to preserve trust.”
Public service media needs to be and usually is ahead of the game on using new tech responsibly. The European Broadcasting Union, whose members are all the main PSM providers in Europe and which runs the Eurovision Song Contest, has been instrumental in setting standards. Last year it published a blueprint on how to harness AI and on “promoting the use of artificial intelligence for the social good.”
As the document said: “AI can significantly support and enhance human creativity by freeing up time, allowing creators to explore new storytelling techniques, increase accessibility and enhance cultural diversity… [But] it must foster rather than diminish the integrity and diversity of human expression.”
EBU IP Senior Legal Counsel Michele Evangelista says when it comes to protecting the copyright of your content, there are certain things you should remember.
One is that a robot crawler is “not allowed to copy content” on your website, “but you cannot actually prevent it from doing so. It’s a flag; it’s not a shield. So basically, you communicate to the crawler that it is not allowed to copy content on the website, but you cannot actually prevent it from doing so.”
However that flag can help you prove the content was taken without consent should the issue progress to court.
Another important element of AI being discussed by the EU, says Evangelista, is the use of Retrieval-Augmented Generation (RAG).
Whereas traditional AI large language models are based on what they have learnt through training (ie reading text and data to accumulate knowledge and then giving answers based on that) RAG directs the LLM to find relevant information from trusted, knowledgeable, pre-determined sources.
It’s kind of like the equivalent of asking an expert university professor a question or [if using a RAG model] asking them the same question but allowing them first to check the most respected, up-to-date sources.
Even if you’re not in the news, current affairs or factual genres it’s worth knowing about these new ways AI are being used to provide answers to people’s questions as what the LLM provides to their queries might be based on or about one of your shows.
As Evangelista explains: “There is not fully-fledged reproduction. It’s not a copy because you don’t engage the reproduction right according to copyright law. So now they’re starting to discuss this issue in the European Union, how to deal with this unauthorised exploitation, which is difficult to categorise into one of the traditional copyrights.”
Takeaways: agreements are evolving over the use of AI performers/presenters; AI models’ new ways of working are happening faster than regulation
Copyright is key
Expect AI in scriptwriting to be on the news agenda.
Authorship in the Age of AI is the title of writer and Writers’ Guild of Great Britain president Jack Thorne’s talk with Baroness Beeban Kidron on Sunday 13 September at the South Bank Centre in which he is taking “an urgent look at what AI means for the future of storytelling, writing and our imaginations.”
His talk will explain that “copyright is not just for lawyers, it is a vital protection against future incursion” and, according to the WGGB, it’s expected that he will argue that “with the cultural industries becoming ever more crucial to the UK’s economy” there needs to be “an urgent look at where we go next” and “what we need from the government and the industry to get there.”
The AI issue is also being raised at the TUC Congress, which starts the same day.
Writers’ Guild of Great Britain General Secretary Ellie Peers said: “We will be talking about how AI is transforming the concept of ‘authorship’ and the most crucial issue of all – copyright.
“Jack Thorne and Baroness Kidron will be flipping the debate from a focus on protecting the assets of those at the top, to discuss how copyright might protect the industry’s entry points for the future - encouraging people to create and tell stories even as AI evolves.”
Takeaways: protections for writers from AI is going to be on the news agenda; ensure you are up to speed on copyright issues
Sponsorship? If your organisation would like to find out about sponsorship and advertising opportunities, email hello@theindiehustle.com