The threat of artificial intelligence to creators, and its value to big tech, is one of the most pressing concerns in the legal and music fields. Current litigation will ultimately shape how AI will be developed and implemented across entertainment with copyright infringement at the forefront of such disputes.
In October 2023, three major music publishers — Universal Music Publishing Group, Concord Music Group and ABKCO — sued Anthropic, an artificial intelligence company. Anthropic builds AI models by gathering information and text from the internet and training the models to produce output based on that internet sourcing, and the publishers alleged it infringed on copyrighted song lyrics.
Related Stories
The publishers alleged that Anthropic infringed on copyrighted song lyrics by copying the text of the lyrics to train Anthropic’s models and allowing their models to generate text that is similar or identical to the copyrighted song lyrics. Specifically, they take issue with Anthropic’s primary product, the Claude chatbot — alleging that when a user asks Claude to generate song lyrics, the chatbot provides “responses that contain all or significant portions of those lyrics.” Because the chatbot is copying those lyrics and distributing them to others, the publishers argue that it is violating one or more exclusive rights of copyright owners.
A market for licensing copyrighted work already exists, the publishers argue, and there are many websites that perform the same function as Anthropic’s AI models that are not infringers of the copyrighted work because those sites have licenses. As such, the plaintiffs allege that without a license to reproduce and distribute the copyrighted song lyrics, Anthropic is depriving both the publishers and the musical composers of the right to earn money from the copyrighted lyrics.
In their complaint, the publishers point out that Anthropic — which has been reported to be valued at $18 billion and to have been funded by Google, Zoom and Salesforce — does not pay anything for the copyrighted lyrics. They emphasize that if Anthropic is allowed to continue using copyrighted material to train its for-profit models, the licenses for which other companies pay to use their copyrighted lyrics will be rendered worthless, as will the market for those licenses.
Anthropic’s ability to garner investment by major tech players — including a recent $4 billion investment from Amazon — reflects an optimism about AI’s potential. Proponents argue that AI’s functions merely imitate human thinking, and as such, Anthropic is likely to employ the fair use doctrine in defending against this lawsuit.
To establish fair use here, Anthropic would need to show that, under these facts, they are permitted to use the copyrighted song lyrics without the plaintiffs’ permission because, among other factors, the output of the chatbot is transformative in some way. Because there are no bright-line rules for arguing fair use, Anthropic would need to present a strong case for its situation — likely involving the development of AI and the need for innovation. Under the existing case law, it is difficult to successfully establish a use is fair where the defendant is commercially benefitting from its use or potentially destroying the market for the copyright owner’s work, as is arguably the case with Anthropic.
AI proponents, including venture capital firm Andreessen Horowitz, have specifically pointed to the billions of dollars invested in U.S.-based AI companies as an essential reason to allow for their continued use of copyrighted materials to train AI models without a license — a cost assumption on which many of these investments have been based. Licensing requirements would involve significant administrative and legal costs for both AI companies and the creatives from which such licenses are sought, which would likely slow the rate at which AI models can be trained. Moreover, the EU has already proposed a regulatory framework for AI that would both grant copyright protection to AI-generated works and require companies to disclose any copyrighted material used to develop their systems, rather than banning such a use outright.
These factors could create large implications for AI generally if Anthropic loses the courtroom battle in Tennessee. A major shift in investments in U.S-based AI companies may follow such a loss.
On Nov. 22, Anthropic filed a motion to dismiss the complaint arguing that Tennessee does not have jurisdiction over Anthropic and therefore that venue is improper. Anthropic argued that it is neither “at home” in Tennessee nor are there specific connections between Anthropic and Tennessee that relate to the complaint. In the alternative, Anthropic argued that the case should be transferred to the Northern District of California, where Anthropic has its principal place of business. Whether or not this case is heard could influence the direction of AI and copyright law generally.
The effects of developing AI are far-reaching, and the legal field is no exception. Music is just one area where creators are bringing suit against AI for its training processes. Photographers and authors, for example, have already been involved in litigation over AI’s use of their work. These lawsuits will shape the boundaries of AI and provide guidance as to what qualifies as copyright infringement — both for training AI models and for what the models produce.
Given that the Anthropic case has the potential to set the precedent for the specific issue of whether copyrighted materials may be used to train AI technologies without a license, it remains to be seen whether a court will determine that these facts constitute fair use under the existing cases (thus declaring that Anthropic is not infringing on the plaintiffs’ copyrights) or whether the extension of the doctrine that far will require further legislation and regulation by Congress and the U.S. Copyright Office.
The U.S. Copyright Office on Aug. 30 began seeking public comments regarding what permission and/or compensation is necessary to use copyrighted material to train AI technologies in order to help assess whether legislative or regulatory steps in this area are warranted. The public comment period offers a critical opportunity for the music business (and other creative industries) to participate as stakeholders in the discourse that will frame the future of AI technologies in the creation of new works, and compensation for those that already exist.
Tatjana Paterno is a member and Layna Deneen is an associate at the law firm of Bass, Berry & Sims in its Nashville, Tennessee office. They counsel private equity funds and corporations on matters including mergers and acquisitions, and capital markets transactions
Emily A. Burrows, Paige Mills and Megan E. Smit also contributed to this column.
