Given the extraordinary pace of development and adoption of generative artificial intelligence ("Generative AI") tools, the Canadian Government, through Innovation, Science and Economic Development Canada, is once again consulting on Canada's copyright framework, this time with a focus on Generative AI. Submissions are made using the Government's online survey, which is open until January 15, 2024 (extended from December 4, 2023).

We referred to this consultation in Tamara Céline Winegust's recent article, "Acquiring and Licensing Content in the Age of Generative AI", which highlights steps rights holders (and prospective rights holders) may wish to take now to best position themselves in the current environment. This post explains the context for the current consultation and provides more detail on the issues that the government is canvassing now.

This new consultation follows a previous consultation held between July 15 and September 17, 2021, "A modern copyright framework for artificial intelligence and the Internet of Things". Since then, rapid proliferation of Generative AI tools, like CHAT-GPT, has led the Government to recognize a "need to reconsider some of the issues studied" two years ago. The new consultation, Consultation on Copyright in the Age of Generative Artificial Intelligence, is intended to focus on the "impacts of recent developments in AI on the creative industries" as well as economic impacts of these technologies, with a view to considering whether change is required to "further improve or reinforce copyright policy" for Canada. The consultation paper highlights three areas of particular interest to the Federal Government: (1) Text and Data Mining (TDM); (2) Authorship and ownership of works generated by AI; and (3) Infringement and liability regarding AI. Each are discussed further below. The consultation launched on October 12, 2023.

What is Generative AI?

Generative AI tools use complex machine learning models that are trained using vast amounts of content and data, including human-authored content such as text, images, and music. Large amounts of Generative AI training materials are "scraped" from websites on the internet. Once trained on such materials, Generative AI tools create outputs, such as text (eg. ChatGPT), images (eg. DALL-E), or music (eg. MusicGen), that correspond to a user's prompt. Generative AI tools have become increasingly sophisticated. Today, the text, images, and music that such tools are capable of generating can be hard to distinguish from something produced by human intelligence.

What issues is the government canvassing now?

(1) Text and Data Mining

Much of the text and data that is mined from the internet for training Generative AI contains "original works" that are subject to protection under the Copyright Act. This creates a tension between the rights of creative industry owners of such copyrights, and the AI industry that wishes to use this content to develop new technologies. The issues under the Copyright Act that arise from this tension are illustrated by the questions that the Consultation Paper canvasses in Section 2.1.3:

i. What would more clarity around copyright and TDM in Canada mean for the AI industry and the creative industry?
ii. Are TDM activities being conducted in Canada? Why is it the case or not?
iii. Are rights holders facing challenges in licensing their works for TDM activities? If so, what is the nature and extent of those challenges?
iv. What kind of copyright licenses for TDM activities are available, and do these licenses meet the needs of those conducting TDM activities?
v. If the Government were to amend the Act to clarify the scope of permissible TDM activities, what should be its scope and safeguards? What would be the expected impact of such an exception on your industry and activities?
vi. Should there be any obligations on AI developers to keep records of or disclose what copyright-protected content was used in the training of AI systems?
vii. What level of remuneration would be appropriate for the use of a given work in TDM activities?
viii. Are there TDM approaches in other jurisdictions that could inform a Canadian consideration of this issue?

(2) Authorship and Ownership

The authorship and ownership provisions of the Copyright Act were drafted with human authors in mind, yet Generative AI tools can produce new creative content based only on a human prompt. The consultation paper asks in Section 2.2.3:

i. Is the uncertainty surrounding authorship or ownership of AI-assisted and AI-generated works and other subject matter impacting the development and adoption of AI technologies? If so, how?
ii. Should the Government propose any clarification or modification of the copyright ownership and authorship regimes in light of AI-assisted or AI-generated works? If so, how?
iii. Are there approaches in other jurisdictions that could inform a Canadian consideration of this issue?

Likely with reference to these questions , the Government has flagged, and is seeking input on, three possible approaches to address the uncertainties surround the authorship and ownership of works generated by AI or created with the assistance of AI:

A. Clarify that copyright protections only apply to works created by humans.
B. Attribute authorship on AI-generated works to the person who arranged for the work to be created.
C. Create a new and unique set of rights for AI-generated works.

(3) Infringement and Liability

As the consultation paper acknowledges in Section 2.3, Generative AI raises novel issues regarding liability for infringement that may result from the use of AI, either through the inputs used to train an AI or through the outputs generated by an AI tool. For example, it could be difficult for a copyright owner alleging infringement in an AI application or AI generated work to identify the person, or persons responsible for direct or secondary infringement, as well as other elements of infringement. The consultation paper asks in Section 2.3.2:

i. Are there concerns about existing legal tests for demonstrating that an AI-generated work infringes copyright (eg. AI-generated works including complete reproductions or a substantial part of the works that were used in TDM, licensed or otherwise)?
ii. What are the barriers to determining whether an AI system accessed or copied a specific copyright-protected content when generating an infringing output?
iii. When commercializing AI applications, what measure are businesses taking to mitigate risks of liability for infringing AI-generated works?
iv. Should there be greater clarity on where liability lies when AI-generated works infringe existing copyright-protected works?
v. Are there approaches in other jurisdictions that could inform a Canadian consideration of this issue?

Interesting Times

In Canada and around the world, it will be interesting to see how the copyright issues being canvassed in this most current consultation are addressed by legislatures and the Courts both in Canada and elsewhere, or if some renewed international effort will be undertaken to harmonize the legal approach to these issues. We are following these issues as they develop to be able to advise our clients on the latest. We will continue to post as the submissions to the consultation are published and as the law develops. Watch this space!

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.