Generative AI and Music: Copyright Risks Highlighted by the DIP

on
an artist s illustration of artificial intelligence ai this image represents the boundaries set in place to secure safe accountable biotechnology it was created by khyati trehan as pa

Generative AI and Music: Copyright Risks Highlighted by the DIP

The growing use of generative artificial intelligence (AI) to create music is raising increasingly important copyright questions. AI tools can now generate songs, modify voices, create remixes and produce new musical content by reference to existing works, making the boundary between technological creation and the use of protected material increasingly significant.

The Department of Intellectual Property (DIP) has recently highlighted the copyright implications of using generative AI in music. While the DIP’s comments do not introduce new legislation or a separate legal regime for AI-generated content, they provide a useful practical signal: the use of AI does not remove the need to consider copyright in the material used as part of the creative process.

For businesses using generative AI for music, advertising and other commercial content, this has implications not only for copyright clearance but also for contracts with AI providers, internal policies and the management of infringement risk.

Existing copyright rules continue to apply:

The starting point is that generative AI does not operate outside the existing copyright framework. Under the Copyright Act, copyright owners have exclusive rights in relation to protected works, subject to applicable limitations and exceptions. Depending on the circumstances, reproducing, adapting or otherwise using a protected work without authorization may therefore constitute infringement.

The DIP has emphasized that where copyrighted material is used in connection with generative AI, users should consider whether they have the necessary rights and obtain permission where required. This is particularly relevant where an AI workflow involves identifiable existing material—for example, where a user supplies an existing song, recording or other protected content to an AI system to generate or modify musical content. The fact that AI technology performs part of the transformation does not, by itself, provide authorization to use the underlying copyrighted work.

AI-assisted music can involve several layers of rights:

Music-related AI applications can be legally complex because a single piece of music may involve multiple protected elements. A song may involve rights in the musical composition and lyrics, while a particular recording may involve separate rights in the sound recording. Depending on how an AI tool is used, more than one category of rights may therefore need to be considered.

For example, using an existing recording as an input for an AI-generated remix may raise different questions from merely instructing an AI system through text to create music of a particular genre. Similarly, an AI voice-conversion tool that processes an existing recording may involve different copyright considerations from a system generating an entirely new recording without the user supplying an existing protected work. Businesses should therefore avoid treating “AI-generated music” as a single legal category. The relevant copyright analysis depends significantly on what material enters the AI workflow, what the system does with that material and how the resulting content is subsequently used.

Copyright clearance should begin with the input:

For businesses, one of the most immediate implications of the DIP’s position is the importance of reviewing the material supplied to AI systems. Before employees, agencies or contractors upload music, recordings or other content to a generative AI platform, businesses should consider whether they own the relevant rights, have obtained an appropriate license or can otherwise lawfully make the intended use.

This is particularly important in advertising and marketing, where AI tools may be used to generate background music, modify existing tracks or rapidly produce multiple versions of creative content. A business may ultimately be responsible for content distributed under its name even where an external advertising agency, production company or AI provider performed much of the underlying creative work. Copyright clearance should therefore form part of the AI-content production process rather than being addressed only after the content has been generated.

AI provider contracts deserve closer scrutiny:

The copyright analysis should not stop with the underlying content. Businesses should also review the contractual terms governing the AI tools they use. Terms of service can differ considerably between platforms, particularly in relation to material uploaded to the platform, the provider’s ability to use customer content and the rights granted in generated outputs.

For commercial use, relevant contractual issues include rights and permissions relating to material submitted to the AI system, permitted use of customer-provided content by the AI provider, rights to use and commercialize generated outputs, intellectual property representations and warranties, indemnification for infringement claims, and procedures for responding to copyright complaints. Similar protections may be appropriate in agreements with advertising agencies, production companies and other contractors creating AI-assisted content.

Internal AI policies should address copyrighted content:

Businesses increasingly permit employees to use generative AI tools without necessarily treating that use as a formal intellectual property process. This can create risk where employees upload commercially released music or other third-party content to an AI platform, use copyrighted material as a reference, or use AI to modify content without considering whether the business has the necessary rights.

Internal AI policies should therefore address intellectual property alongside confidentiality, personal data and cybersecurity concerns. Organizations should consider establishing rules governing the types of third-party content that may be uploaded to AI systems, when copyright clearance is required and which AI platforms may be used for commercial content creation. For higher-risk uses, an internal approval process may also be appropriate before AI-generated material is released publicly or incorporated into a commercial campaign.

What the DIP’s position does—and does not—resolve:

The significance of the DIP’s comments should not be overstated. They provide a useful indication of how existing copyright principles should be approached when generative AI is used to create or modify music and reinforce the practical importance of obtaining authorization before using copyrighted works where permission is required.

However, the comments should not, without further legal or regulatory authority, be treated as establishing a definitive position on whether and under what circumstances copyrighted works may be used to train generative AI models. Nor should they be treated as conclusively determining whether, or under what circumstances, AI-generated output qualifies for copyright protection or who may own rights in such output. Those questions involve distinct legal issues concerning reproduction, exceptions to copyright, authorship, originality and the degree of human creative contribution.

Practical implications for businesses:

Companies using generative AI to create music or other commercial content should consider incorporating copyright review into their AI governance framework. A risk-based approach may be appropriate: generating content from text instructions without supplying identifiable third-party works may present a different risk profile from uploading existing songs or recordings, generating remixes or adaptations, or using protected material as a direct input or reference in the generation process.

Particular caution is appropriate where AI-generated content will be used in advertising, distributed commercially or incorporated into products. Businesses should also consider the complete contractual chain. An organization commissioning AI-generated music from an agency or contractor may wish to require appropriate warranties concerning the lawful use of source material rather than assuming that copyright compliance rests exclusively with the creator.

Key Takeaways:

  • Generative AI does not displace copyright law: Using an AI tool does not, by itself, authorize the reproduction, adaptation or other use of copyrighted material.
  • Inputs matter: Businesses should understand what copyrighted material is being supplied to an AI system and whether the necessary rights or permissions have been obtained.
  • Music can involve multiple rights: Compositions, lyrics and sound recordings may involve separate rights and require separate analysis.
  • Contracts should allocate AI-related copyright risk: Businesses should review AI-provider and agency agreements for input rights, output rights, warranties, indemnities and restrictions on the provider’s use of uploaded material.
  • Internal AI policies should cover intellectual property: Rules governing employee use of generative AI should address copyrighted inputs and commercial use of AI-generated content.
  • Important questions remain unresolved: The DIP’s comments should not be interpreted more broadly than their stated scope, particularly regarding AI training and copyright ownership of AI-generated output.

Author: Panisa Suwanmatajarn, Managing Partner.

Other Articles

Posted in

Tagged as