The US Supreme Court has said it will not intervene in a key legal battle over the US Copyright Office’s view that AI-generated works are not protected by copyright under American law. The Supreme Court’s refusal to review the case means the Copyright Office’s position on the copyrightability of AI-generated content stands, for now at least.
Physicist and inventor Stephen Thaler has been seeking copyright protection for AI-generated works since 2018, before generative AI became such a major talking point across the music and wider creative industries.
At the centre of the case is a piece of visual art that was generated by an AI model called ‘DABUS’ that Thaler developed, which he says “conceives new inventions and artforms”.
He first tried to register his image, called ‘A Recent Entrance To Paradise, with the Copyright Office in 2018 and, after it rejected his registration in 2022, he began his legal battle through the courts. Which have all backed the Copyright Office’s position, something Thaler wanted the Supreme Court to overturn.
Although this case centres on a piece of visual art, the principle being considered applies to all creative works, including music, which is why Thaler’s lawyers told the Supreme Court this dispute was of “paramount importance” in a world where so many people are now using AI to generate content.
Responding to the Supreme Court’s decision, Thaler’s legal team told Reuters that - even if the top court considers a similar case in the future and overturns the Copyright Office’s position on this - “it will be too late, the Copyright Office will have irreversibly and negatively impacted AI development and use in the creative industry during critically important years”.
It is generally assumed that, under most copyright systems around the world, entirely AI-generated works do not enjoy copyright protection. Which means anyone can make use of those works in any way without getting the permission of anyone else. That’s based on the principle that - to quote one of the judges who ruled on the Thaler v Copyright Office dispute - human authorship is a “bedrock requirement of copyright”.
Because in the US there is a copyright registration system, it was easier to test that assumption, as Thaler did by registering an entirely AI-generated work and seeing if the Copyright Office would accept it.
Which it did not, of course. More recently the Copyright Office published a report on the copyrightability of AI-generated works, concluding that if a human uses AI as part of the creative process their work will be protected by copyright, but entirely AI-generated works have no such protection.
It then had a go at identifying how much human involvement there needs to be to say the AI only assisted in the creative process, concluding that “prompts alone” are not enough.
There are, of course, two aspects to this debate. First, does copyright law, as it is currently written, provide protection for AI-generated works. But second, should it?
Because the law can always be changed if people agree with Thaler that AI-generated works should enjoy the same copyright protection as human created works.
Within the music community, you sense most people believe AI-generated works should not have copyright protection. Though opinions may differ between different stakeholders, and over time, because it’s an issue where the idiom “be careful what you wish for” may be relevant.
If AI-generated works are not protected by copyright, and therefore readily available to everyone free of charge, does that mean catalogues of AI-generated music pose a bigger threat to human creators?
For streaming services and AI music start-ups like Udio and Suno, there are arguments that these companies benefit from AI works having copyright protection, but also if they do not. It may be that the position these companies take changes as their business models evolve.
UK copyright law is unusual, in that it arguably does provide copyright protection for AI-generated works, because the UK Copyright Act sets out ownership rules for what it calls ‘computer-generated’ works.
Whether or not that is the right approach was one of the questions posed by the UK government when it opened up a consultation on copyright and AI at the end of 2024.
The government’s consultation document wrote, “the UK currently provides copyright protection for purely computer-generated works, but it is not clear that this protection is widely used, or that it functions properly within the broader copyright framework”. As a result, “the government seeks views on potential reform to protections for computer-generated works”.
Ministers are due to publish a plan on copyright and AI based on that consultation later this month, and we’ll see what, if anything, they say regarding whether or not AI-generated works should enjoy the protection of copyright. But even if they do under the UK system, for now at least, they are not protected in the US. And the Supreme Court isn’t going to intervene on that anytime soon.