The Rise of AI-Generated Music: Legal Considerations
New technology is sure to bring three things along with it: Fear, excitement, and uncertainty. Artificial Intelligence (“AI”) is no different. But what may be different about AI is the speed at which it is developing and accordingly, raising a plethora of legal issues.
This is especially true with the rise of AI-generated music, and the resulting copyright issues, along with right of publicity considerations. In addition to the actual AI software used, there are several moving parts when it comes to AI-generated music, content, and rights, the most prominent involving: (1) The content used to train AI software (the so-called “source material”); and (2) The content used to create and market the AI-generated music (e.g., musical compositions, sound recordings, and artists’ likenesses and/or voices).
While this issue finally got the attention of Congress, as evidenced by a Congressional hearing held on May 16, 2023 with the OpenAI CEO1, existing law has yet to specifically catch up with the scenarios and issues that are unfolding every day with the explosion of AI-generated music. As the law develops, there is no choice but to work within the existing intellectual property and right of publicity frameworks. That means applicable parties that want to release or otherwise exploit AI-generated music obtaining the proper clearances.
At the heart of copyright law is the balance between encouraging innovation and protecting creators’ rights. While AI developers may plan to argue that their uses of source material constitute fair use under copyright law2, it is unclear whether they would be successful with such argument. AI music creators also play a role and tread the sensitive line between innovation and the rights of the stakeholders in the source material and/or rights of publicity that AI music creators may use.
Rights of publicity are based on state law and may often have limited protection or barriers to enforcement, such as first amendment rights potentially applying to protect would-be defendants in certain situations.3 However, such rights are currently the basis of recent talks between major labels and digital music providers to potentially find a procedure by which to takedown AI-generated music where copyright law may fall short.4
Given the current state of the law, industry concern, and basic ethical considerations, being proactive seems essential in limiting risk and overall, “doing the right thing”. This means AI developers potentially obtaining licenses for specific source material before using such material and along the same lines, AI music creators obtaining licenses for specific content before using such content. Depending on the type of content AI developers and/or music creators plan to use and how they plan to use it, these could involve licenses for musical compositions, sound recordings, and various rights of publicity.
If you are an AI developer or music creator who wants help with licensing, or simply have questions about AI-generated music, it only makes sense to have knowledgeable and experienced lawyers on your team. Contact us today to learn more about potentially working together.
By: Neiloofar Sajedi, Esq.
DISCLAIMER: In addition to the disclaimer, for clarity: This information is for informational purposes only and should not be taken as legal advice for any individual case or situation; and please do not consider this information to be a substitute for obtaining legal advice from a qualified attorney licensed in the jurisdiction(s) relevant to your matter.
Sources:
Oversight of A.I.: Rules for Artificial Intelligence: Hearing Before the Subcomm. On Privacy, Technology, and the Law, 118th Cong. (2023).
17 U.S. Code § 107.
Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989); Mattel, Inc. v. MCA Records, 296 F.3d 894 (9th Cir. 2002).
Kristen Robinson, Major Labels to Send ‘Takedown Notices’ to Streaming Services for AI Soundalikes, BILLBOARD (May 16, 2023), https://www.billboard.com/pro/major-labels-streaming-services-take-down-ai-soundalikes/#recipient_hashed=abf57c7cb995fc29334418b56daa22d936067bc2361dc686efb7020fac565776&recipient_salt=702cfa69270844f6c812632a5ec650382b5e882b7cee94386bfb000e8b3205c6.