AI in music: The Drake/Tupac case and beyond

7 mai 2024 | Kurt Dahl, David Krebs, Alissa Ricioppo

( Disponible en anglais seulement )

In a matter of days, Drake’s latest release went from making headlines around the world to being taken down amidst threats of legal action from the estate of arguably one of the most revered and popular rap artists of all time, Tupac Shakur. The crux of the matter: the struggle between copyright and the use of artificial intelligence (AI) in music. The case brings up other legal issues such as the right of personality and privacy rights, which we will discuss below.

The Drake/Tupac controversy

Drake’s song, « Taylor Made Freestyle », was released on April 19, 2024 and a few days later the pop star received a cease-and-desist letter from Tupac Shakur’s estate over the song’s use of AI. The Shakur estate called it « a blatant abuse of the legacy of one of the greatest hip-hop artists of all time. »

The source of this “blatant abuse” is the prominent use of AI-generated voices that bear an indistinguishable resemblance to both Shakur and another famous rap artist, Snoop Dogg. The similarity between the AI-generated voices and the real artists is uncanny, which reflects the level of sophistication and power of current AI technologies.

The song is the latest to draw attention and controversy for its use of AI. Less than a year ago, an AI-generated likeness of Drake’s voice was used in a popular track called “Heart on My Sleeve” by producer Ghostwriter977. Universal Music Group (UMG) had the track removed from all platforms, proffering that such use of AI “demonstrate[s] why platforms have a fundamental legal and ethical responsibility to prevent the use of their services in ways that harm artists”.

How does AI-generated music work?

AI-generated music involves the use of algorithms and machine learning techniques to create new compositions or recordings. These algorithms ingest and then analyze vast amounts of existing music data (which includes the voices of the any vocalists involved) to identify patterns, styles, and structures, which they then use to generate new creations. From classical symphonies to contemporary pop tunes, AI can mimic various genres and even produce compositions that resonate emotionally with listeners.

Blurred lines between AI and copyright

In the realm of song writing and recording, the lines between human ingenuity and AI innovation continue to blur. With advancements in AI technology, technology is now capable of generating melodies, harmonies, and entire compositions. However, as the boundaries of AI-generated music expand, so too do the questions surrounding its legality and ethics.

The copyright conundrum

The legality of AI-generated music hinges largely on copyright law. Copyright protects original works of authorship, including musical compositions. However, the question arises: who owns the copyright when AI creates music?

In most jurisdictions, copyright law attributes authorship to human creators. Since AI lacks legal personhood, it cannot be considered the author of a creative work. Therefore, the copyright for AI-generated music typically defaults to the individual or organization that owns and operates the AI system. This could be the programmer, the entity that trained the AI model, or the user who initiated the music generation process.

Challenges and controversies

The rise of AI-generated music poses several challenges and controversies within the legal landscape. One primary concern is the potential infringement of existing copyrights. AI algorithms, when trained on copyrighted music, may inadvertently produce compositions that resemble existing works too closely, raising issues of plagiarism and intellectual property infringement.

Additionally, there is the question of fair compensation for human creators. If AI-generated music becomes widespread in commercial settings, musicians, composers, and songwriters may see their livelihoods threatened. Ensuring fair compensation and recognition for human creativity amidst the proliferation of AI-generated content is a crucial aspect of navigating the legal terrain of music composition.

A tale of two recordings

With “Heart on My Sleeve,” the recording was touted as a new release featuring both Drake and the Weeknd, and it instantaneously went viral, becoming the most talked about song on the internet. Interestingly, UMG was successful at having it taken down not because of the use of the “cloned” voices of the two superstars, but because the recording contained a musical sample which was not cleared or authorized, resulting in an easy takedown for UMG. The question as to whether Ghostwriter977 could use the “cloned” voices remained unanswered.

With « Taylor Made Freestyle, » the musical composition was presumably original and the recording presumably contained no unauthorized samples, leaving only the cloned voices of Tupac and Snoop Dogg at issue.

This begs the question: does the use of “cloned” vocals constitute a violation of copyright? Can an artist copyright or protect a style or sound of their vocal? Can our current copyright framework deal with this question or is it something else entirely? The song was taken down by Drake but no reason was given. Was this a reaction to the pressure exerted by the Shakur estate? Or did his legal team advise him that the recording was a violation of the rights of Shakur and Snoop Dogg?

Right of personality

Beyond copyright, the use of “cloned” vocals raises issues related to personality rights. Personality rights refer to one’s right to control the use of their own name, image, voice and other unique identifying features.

In cases where one’s personality is exploited for commercial purposes and without their permission, as was the case here with Shakur and Snoop Dogg’s vocals, they may have a claim for what is called misappropriation of personality. This type of claim is recognized under Canada’s common law. Most misappropriation of personality cases involve famous plaintiffs, but some courts have stated that fame is not a requirement for a successful claim.

Importantly for cases like this one, courts have confirmed that personality rights may be asserted by an estate following the death of the individual at issue, though the law is unclear as to how long the rights remain in play after the subject’s death.

Privacy issues

In some provinces, like Saskatchewan, British Columbia, Newfoundland and Labrador and Manitoba, there is legislation in place that recognizes misappropriation of personality as a wrongful act. In some provinces, it is treated as a form of a violation of privacy. Unlike under the common law, some of these provincial laws specifically state that personality rights extinguish on death.

One must also remember that a voice print is “personal information” under Canadian and many other foreign privacy laws. The collection of personal information for use in training AI models would be subject to those privacy laws and to potential consent requirements, which raises a host of potential issues for the use of AI technologies trained on data sets that may not have been obtained compliantly.

Legal precedents and future perspectives

The Drake and Tupac case is just one example of the legal challenges posed by AI-generated music. As the technology continues to evolve, legal frameworks must adapt to address emerging challenges. Some jurisdictions have already taken steps to clarify the legal status of AI-generated creations. For example, in 2020, the European Union Intellectual Property Office ruled that an AI system cannot be listed as an inventor on a patent application, emphasizing the need for human authorship in intellectual property rights.

Looking ahead, legislators, policymakers, and legal scholars will need to engage in ongoing dialogue to develop comprehensive regulations and guidelines for AI-generated music. These frameworks should balance the promotion of innovation and creativity with the protection of artists’ rights and interests. Clear guidelines on ownership, licensing, attribution, and royalties will be essential to foster a vibrant and equitable ecosystem for music creation in the age of AI.

Conclusion

The use of AI to write music raises complex legal questions regarding copyright, authorship, and intellectual property as well as personality and privacy rights. While AI-generated music offers exciting possibilities for creativity and innovation, it also presents challenges in terms of legal clarity and ethical considerations. As technology continues to advance, the music and tech industries must collaborate to establish robust legal frameworks that promote innovation while upholding the rights and interests of human creators.

By navigating these challenges thoughtfully, we can unlock the full potential of AI-generated music while ensuring a harmonious and equitable future for all stakeholders in the music industry.

If you would like further information or to discuss these issues, please reach out to a member of our Technology, Intellectual Property, and Privacy Group

Avis de non-responsabilité

Cette publication est fournie à titre informatif uniquement. Elle peut contenir des éléments provenant d’autres sources et nous ne garantissons pas son exactitude. Cette publication n’est ni un avis ni un conseil juridique.

Miller Thomson S.E.N.C.R.L., s.r.l. utilise vos coordonnées dans le but de vous envoyer des communications électroniques portant sur des questions juridiques, des séminaires ou des événements susceptibles de vous intéresser. Si vous avez des questions concernant nos pratiques d’information ou nos obligations en vertu de la Loi canadienne anti-pourriel, veuillez faire parvenir un courriel à privacy@millerthomson.com.

© 2024 Miller Thomson S.E.N.C.R.L., s.r.l. Cette publication peut être reproduite et distribuée intégralement sous réserve qu’aucune modification n’y soit apportée, que ce soit dans sa forme ou son contenu. Toute autre forme de reproduction ou de distribution nécessite le consentement écrit préalable de Miller Thomson S.E.N.C.R.L., s.r.l. qui peut être obtenu en faisant parvenir un courriel à newsletters@millerthomson.com.