We use cookies to collect and analyse information on site performance and usage to improve and customise your experience, where applicable. View our Cookies Policy. Click Accept and continue to use our website or Manage to review and update your preferences.


Music rightsholders vary in approach to AI scraping
Robert Plant and Jimmy Page of Led Zeppelin perform at Nassau Coliseum on February 13, 1975 Pic: Shutterstock

03 Sep 2024 IP Print

Music rightsholders vary in approach to AI scraping

ALG lawyers have said that technological advances in generative AI pose challenges from a legal perspective, particularly in relation to how they may impact the intellectual-property rights of musicians.

Generative AI tools can compose based on input parameters such as genre, tempo, mood, and training data, the lawyers state

But these require vast amounts of training data, raising concerns for rightsholders whose works may be used in training these models, risking copyright infringement, write ALG solicitors Aideen Burke, Eimear McCluskey and Shannon Owens.

A single song may have many rightsholders (writer, performer, publisher, record label,) and clearing use for AI training by way of licensing will require a number of different licences for the same song.

Partnerships between technology companies and rightsholders may emerge for use of existing musical works as AI training data.

This could potentially be in the form of a blanket licensing regime being adopted by key music-collecting societies, the ALG lawyers state.

Unauthorised AI-generated content is now being addressed by online platforms such as Facebook, Instagram, Messenger, Horizon, Threads, and WhatsApp.

Further collaborations between publishers or labels and online platforms will emerge, with a more formalised approach to the re-use of existing music to create AI-generated content.

Exceptions

Statutory copyright exceptions apply for research for non-commercial purposes where this use is sufficiently acknowledged.

This is covered under article 3 of the EU Directive 2019/790 on copyright and related rights in the Digital Single Market (EU Copyright Directive).

This exception for non-commercial research is unlikely to be useful for training AI models using copyright-protected works, ALG states.

Most companies developing large generative AI models intend to commercialise their outputs.

A person can also reproduce a copyright-protected work where the author has not expressly reserved the use of the work for text and data mining (TDM).

The EU Artificial Intelligence Act requires compliance with EU law on copyright and related rights, especially reservations of rights under article 4(3) of the EU Copyright Directive.

This obligation will apply 12 months after the AI Act comes into force.

Sony Music Group (SMG) has already prohibited and opted out of any text or data mining, web scraping or similar reproductions, extractions or uses of any of its content for any purpose without its authorisation.

SMG has also reportedly directly warned hundreds of technology companies not to use its music to train AI.

An Irish Music Rights Organisation (IMRO) statement of 5 June gave a similar warning.

Fair dealing

Irish copyright law provides for a broad ‘fair dealing’ exception, which permits use of copyright-protected works for research or private study, caricature, parody or pastiche, criticism, and review.

This narrow exception is unlikely to be helpful to most AI technology companies.

Under Irish law, it is not an infringement of copyright to undertake or conduct temporary acts of reproduction that are transient or incidental and are an integral and essential part of a technological process.

Under Irish copyright law, the author of a work must be a natural or legal person (not a machine).

Navigating copyright issues for AI-generated music will be increasingly complex, the ALG lawyers state.

While some rightsholders are clamping down hard on copyright infringement with legal proceedings, others are seeking strategic licensing arrangements with technology companies.

AI developers must ensure that the methods they adopt to train their AI models are compliant with applicable copyright law, and their own policies as required under the AI Act.

They must respect express reservations by rightsholders (such as Sony and IMRO) in respect of the use of their works for the purposes of TDM, or face potential copyright-infringement claims.

Ownership of copyright in works created using complex generative AI tools is yet to be tested before the Irish courts, and it remains to be seen whether it is possible for copyright to subsist in such works, the ALG intellectual-property lawyers state.

Gazette Desk
Gazette.ie is the daily legal news site of the Law Society of Ireland

Copyright © 2024 Law Society Gazette. The Law Society is not responsible for the content of external sites – see our Privacy Policy.