Let’s make one thing clear. This article was not written with the aid of ChatGPT. Although perhaps that is exactly what a truly intelligent artificial intelligence (AI) would say. The risk of AI doing sneaky things like that (and far worse) has of course stimulated Elon Musk and several hundred AI experts to sign an open letter calling for an immediate suspension of next-generation AI development and the “dangerous race to ever-larger unpredictable black-box models.” They argue such a pause is needed to “develop and implement a set of shared safety protocols” and allow “policymakers to dramatically accelerate development of robust AI governance systems.” Which does make it sound like a T-800 with an uncanny resemblance to an Austrian bodybuilder turned politician is being fine-tuned in a lab near you right now.
Even if the Terminator is not imminently expected in our lives, the sudden availability–and ability–of AI systems like ChatGPT has certainly been a wake-up call for many, even replacing discussion of NFTs and the Metaverse as the rights-owner technology quandary of the day. As an opportunity and a threat. That’s not to say that many in content businesses, and publishing in particular, haven’t already had to think hard about the impact of AI and machine learning. Text and data mining (TDM)–made powerful through the use of computer algorithms–and the scope of statutory and/or contractual permissions for that, which have been a topic of debate for many years. As have the risks (and required retractions) of fake research papers and so-called “paper mills” magnified by the application of AI, compromising integrity in STM publishing.
Further, in the case of less academic publishing content, the ability of social media platforms to curate, target, and exploit user-generated content with significant success through targeting and personalization is driven by sophisticated software systems and not the inherent quality of the underlying content. A truism which is not lost on owners of more professional material competing for the same user eyeballs, spend, and engagement time, and increasingly adopted on the platforms publishing that material. Moving beyond the presentation and user experience of published content, there is also a belief that AI will be able to develop and create that content–taking the traditional gamble out of spotting the next bestselling author, indeed potentially replacing the bestselling author. So will the Terminator need an agent?
That’s where the lawyers and law makers have already been busy, looking beyond existing TDM and fair dealing/use exceptions at how generative AI systems drawing upon the works of human creators, and the valuable commercial interests behind those works, can and should be regulated. Relevant U.K. Government stakeholders have produced a range of consultations and responses in recent years grappling with whether, and how far and how quickly, legal reform is needed. Actually, the U.K. is one of the few countries with an existing copyright statute which has long recognized computer-generated works (for more than 30 years), although reported case law would suggest the relevant provisions have not been widely applied or enforced during that time. And the consultations have questioned if AI works should still be protected in that way.
Even if our bionic bestseller does receive copyright protection, where does that leave the copyrights of the human authors, and their publishers, that the AI system may have drawn upon to create it? Not to mention their relevant moral, privacy, personality, and trademark rights. Infringement litigation has only just begun, and content businesses of all kinds are rushing to develop internal and external policies putting corporate guard rails in place in an attempt to manage the risks. Such policies also need to take account of wider AI regulation. Laws (not merely guidance) are now emerging around the world–and there are many dozens of them, with the E.U. in particular pushing ahead with several new Regulations and Directives. Notwithstanding the concerns expressed by the experts’ letter, the legal race has also begun. Regulation will be back.