As artificial intelligence challenges and reshapes the creative ecosystem, SOCAN stands firm on a simple truth: human creativity must be protected, respected, and fairly compensated.
Artificial Intelligence
Generally speaking, Artificial intelligence (AI) refers to computer systems that are designed to perform tasks that normally require human intelligence. These tasks include things like learning, reasoning, recognizing patterns, understanding language, making decisions, and even creating new content.
Instead of just following fixed instructions, AI systems can analyze data, adapt, and improve their performance over time.
AI brings many positive changes to music.
AI has long assisted with mixing and mastering and can restore old recordings with remarkable clarity. For listeners, AI powers personalized recommendations, helping fans discover new artists and genres they might not have found otherwise. In short, AI can act as both a creative partner and a powerful assistant for music creators and fans alike.
AI has also introduced significant challenges in music.
Questions about copyright and ownership arise when AI-generated systems are trained on existing songs, often without music creators’ consent. There’s concern that careers in music creation will be displaced with the flood of AI-generated works in the marketplace. “Deepfake” vocals, which are AI-generated vocals that mimic the vocals of real singers’ voices, also raise both copyright and ethical issues.
These challenges highlight the tension between innovation and the preservation of human creation in music.
The use of music by generative AI platforms
Generative AI is a subset of AI and a powerful new class of tools that can produce complete songs – music and lyrics – in response to a simple written prompt or series of prompts.
These tools are distinct from long-standing studio technologies used for sound recording, editing, and mastering processes because they can create entirely new musical outputs based on patterns learned from massive datasets. Massive datasets that are currently built using the music of SOCAN members and music creators around the world.
SOCAN recognizes the creative potential of generative AI, but we also know it raises real risks for the livelihoods of composers, songwriters and publishers.
Our goal: ensure AI develops in a way that respects, credits and compensates human creators and preserves a healthy market for music.
Navigating AI in Music: Principles and Direction
SOCAN believes generative AI, when used ethically and with a human-centric approach, can be an exciting creative tool, however, it is important that we ensure it credits and compensates human authorship.
Without the right safeguards and respect for the existing copyright framework, it will undermine the incomes and rights of music creators, and the culture of music creation in Canada.
We advocate for:
- copyright and AI policies that foster and reward human creativity;
- transparency and record-keeping requirements for AI developers;
- the ability for creators and their representatives to control how works are used by whom, and on what terms.
Licensing is possible, practical, and necessary to uphold the principles of copyright. As the largest and most active licensor of musical works in Canada, SOCAN firmly believes that licensing musical works to AI companies is not only achievable and practical, but it is the best – and most appropriate – way to support the fundamental purpose of copyright.
SOCAN AI principles respecting human creation
Building the Road Ahead
SOCAN is taking several steps to protect our members’ rights and musical works, such as:
- Advocacy — we lobby the Canadian government for generative AI rules that mandate transparency, support licensing markets and protect creators’ incomes.
- Licensing & contract safeguards — we reserve SOCAN’s existing and future rights relating to AI in our licensing agreements to preserve members’ options and revenues.
- Member education — regular webinars, guidance and resources to explain AI, and provide insights and information on how these systems interact with copyright rules.
- Collaboration — working with international collecting societies, industry groups and governments to shape consistent, practical rules.
SOCAN is no stranger to technological developments. We have a long history of finding solutions that compensate our members for their hard work. We believe generative AI will be no different.
Frequently asked questions
Could generative AI platforms have trained on my music without permission?
Yes, it is possible. Many companies have not publicly disclosed their training data sets.
Can I identify AI outputs that use my songs?
Not reliably today. That’s why SOCAN is pressing for transparency and record-keeping requirements, to ensure rightsholders know when and where their work is being used and use that information to either stop the activity or be compensated for it.
Can I stop or opt out of having my songs being used to train models?
Generative AI developers who use musical works without the permission of the copyright owners expose themselves to potential liability for copyright infringement. There are several ongoing lawsuits in Canada and internationally on this issue.
The “opt-out right” refers to a copyright exception that an author can opt-out of. There is currently no exception for generative AI that an author can “opt out” of under Canadian copyright law. The default rule is that if someone wants to use the work of a music creator in Canada, they must have the copyright owner’s permission to do so.
In contrast, in the European Union there is an exception for the use of lawfully accessible works and other subject matter for the purposes of text and data mining (this is the term used for generative AI companies to train their models). The exception does not apply where the copyright owners reserve, or “opt out”, their rights from being used for this purpose. Many EU societies have expressly “opted out” their members’ rights from this exception.
Are AI platforms breaking the law if they train on copyrighted works without permission?
Unauthorized use of copyright-protected music can expose developers to infringement claims. There are active legal cases worldwide testing these questions.
Are there AI regulations in Canada now?
Currently, Canada has no specific generative AI regulation dedicated to music. Existing copyright law continues to apply. So, if a generative AI developer is infringing copyright, they could be liable for copyright infringement. Other jurisdictions are also actively working on this. The E.U. has already established a regulatory framework (EU AI Act), while the U.S. continues to develop their regulatory framework for generative AI.
What is the position of our federal government?
The Canadian federal government has conducted two consultations on generative AI and copyright and have heard from a variety of stakeholders. SOCAN participated in both consultations and has had meetings with government officials to lobby for regulations that protect and support creators. The Canadian federal government proposed a draft bill to govern generative AI before Parliament was dissolved for the 2025 federal election. To be implemented, Canadian generative AI legislation would have to be introduced again with the new federal government.
How could generative AI affect my income?
Independent studies and industry organizations have warned that, without regulatory safeguards and licensing frameworks, creators could face significant income displacement. SOCAN is pushing for policy solutions to prevent such losses.
Resources & further reading