Subscribe to get latest news delivered straight to your inbox


    UK Parliamentary Committee Shoots Down Copyright Exemption for AI Developers–But is it Really Dead?

    • 24.09.2023
    • By Hugh Stephens Blog
    Hugh Stephens Blog

    The all-party Culture, Media and Sport Committee of the British House of Commons has just (August 30) issued a welcome report condemning plans announced by Britain’s Intellectual Property Office (IPO) in June of 2022 to create a limitless copyright exemption to enable tech companies to develop artificial intelligence (AI) applications without having to obtain permission from rights holders or to acquire a licence. I commented on the IPO’s recommendations in a blog post back in November (“Britain’s Proposed Approach to Text and Data Mining (TDM) for AI: How Not to do It (A Lesson for Canada and Others”), calling it “nothing short of outrageous”. The UKIPO’s proposal meant throwing the British cultural community under the bus in a misguided attempt to make Britain more attractive than the EU for AI developers by making the British Text and Data Mining (TDM) exception broader and less restrictive than the one that exists currently in EU law (which allows TDM only for non-commercial research purposes).  All this supposedly in the name of promoting innovation. To me this reeked of one more desperate attempt to show that Brexit was not the colossal mistake that it clearly was. The IPO claimed (at para. 60) “These changes make the most of the greater flexibilities following Brexit. They will help make the UK more competitive as a location for firms doing data mining.”

    It would be an understatement to say that the IPO’s announcement that the government had “decided” to introduce a new copyright and database right exception to allow permissionless TDM for any purpose set off a furor. It was vigorously attacked by various British cultural industries, from publishing to music to film-making. The CEO of UK Music, Jamie Njoku-Goodwin, called the IPO’s proposals “a green light for music laundering”. While boosters of Britain as an innovation hub played up all the “jobs of the future” that would (maybe) be created, they seem to have forgotten that cultural industries contributed £115 billion in 2021 to the British economy, employing over 2 million people while being a significant source of exports as well as “softpower”. In the face of the furious reaction from many quarters (which, frankly, the government should have anticipated), earlier this year it was finally announced that the proposal had been shelved.  The IPO’s ill-conceived proposal was attacked not just by cultural industry groups and creators, but also by Parliamentarians. The first to take aim was the House of Lords Communications and Digital Committee, now followed by the Culture, Media and Sport (CMS) Committee of the House.

    The main findings of the CMS Committee were pointed and scathing;

    The Committee is pleased that the Government has been listening to stakeholders on text and data mining intellectual property for commercial benefit and is encouraged that Ministers are looking again at this…

    The Government must work to regain the trust of the creative industries following its abortive attempt to introduce a broad text and data mining exemption. The Government should consider how creatives can ensure transparency and, if necessary, recourse and redress if they suspect that AI developers are wrongfully using their works in AI development…

    The Government’s initial handing of the text and data mining exemption to copyright for AI development, though eventually correct, shows a clear lack of understanding of the needs of the UK’s creative industries. All branches of Government need to better understand the impact of AI, and technology more broadly, on the creative industries and be able to defend their interests consistently.

    Apart from criticizing the earlier approach, the Committee strongly supported the status quo with regard to TDM,

    “…the current framework, which provides an exemption for text and data mining for non-commercial research purposes and otherwise allows creators to licence their work for any further purpose, provides an appropriate balance between innovation and creator rights”. (emphasis added)

    It also made other concrete recommendations such as addressing skills and personnel shortages in both the hi-tech and creative sectors, improving digital skills, and strengthening protection for creators to prevent misuse of their likenesses and performances by emerging technologies. (These are some of the same concerns currently at issue with the Hollywood writers and actors strike).

    This is all good stuff, but will it stick? Don’t count out the power of Silicon Valley and the tech lobby in Britain. While the Culture, Media and Sport Committee has done good work through its research and hearings, and has made sensible recommendations, remember that it is just one all-party Parliamentary Committee among many. While chaired by a Conservative MP, the government is under no obligation to take its recommendations on board. True, for the moment, the British government has backed away from the “decision” announced by the IPO, but various proposals are still under study. Not insignificant is the fact that Prime Minister Rishi Sunak is planning on holding the world’s first AI summit before the end of this year, at Bletchley Park, of Enigma fame. (I remember well my visit to Bletchley a couple of years ago. It helped bring all the books I’ve read about the “codebreakers”, of which there are many—books that is–to life). The focus will be on AI safety, but it is part of Sunak’s agenda (remember, he has a high-tech business background) to make Britain a leader in AI. If copyright is seen to get in the way of this ambition, the fight will not be over. That is probably why UK publishers have lobbied Sunak to make it clear that IP rights must be respected when any content is ingested by AI systems by adding this topic to the agenda of the AI summit.

    It may also be telling that in the middle of all the debate about whether AI developers should be given the keys to the copyright kingdom, at no charge, the former Department of Digital, Culture, Media and Sport (DCMS) had the digital file taken away from it and conferred on a newly created department, the Department for Science, Innovation and Technology (DSIT). And let’s not forget that the Intellectual Property Office (responsible for the Copyright Act, among others) does not report to the minister responsible for the Department of Culture, Media and Sport, nor to its predecessor DCMS, but falls under the ministerial responsibilities of DSIT. This is reminiscent of Canada where statutory authority for the Copyright Act, and for the Canadian Intellectual Property Office which administers the Act, falls under the Department of Industry, currently labelled Innovation, Science and Economic Development (ISED). While the “culture” ministry, Canadian Heritage, plays an important role in shaping copyright, the lead rests with ISED where, in my opinion, it gets short shrift. Given that the UKIPO, responsible for copyright, was the author of the unprecedented copyright exemption proposal, one cannot help but wonder when and how Silicon Valley’s acolytes will strike back.

    The report of the Culture, Media and Sport Committee is important and welcome, but don’t expect this issue of “free” access to copyrighted content, wrapped up in the garb of “innovation” (in the US the operative phrase would be “transformation”), to go away. Creators must remain alert while working to find ways to make it possible for AI developers to access the copyrighted content they need to produce credible outputs. Already discussions are underway between rights-holders such as media companies and image/photostock management companies such as Shutterstock, and AI companies, to provide access to valuable, curated content. AI developers can scrape unlimited garbage off the internet, and they do, but if AI generated works are to have credibility and contribute to learning, they need to ingest reputable content and respect the rights of those who have created it. Furthermore, the outputs have to be legal. What is, and is not, legal at the moment is very much open to question. Ingestion of licensed content would help make the case for legality (non-infringement) of AI generated content.

    Giving AI developers a blanket free pass, as the UKIPO originally suggested, is clearly not the way to go. The Culture, Media and Sport Committee report has helped redress the balance. But “it ain’t over ‘till its over”. Creators in Britain and elsewhere need to stay alert, speak up and fight to protect their intellectual property or the hi-tech industry will take another free ride.

    This article was first published on Hugh Stephens Blog