“What We Heard”: Government of Canada Releases Report On Copyright and Generative AI Consultation

On February 11, 2025, the Government of Canada (the Government) published its report summarizing the feedback received from a recent consultation regarding the effects of generative artificial intelligence (AI) on copyright policy and the marketplace. Overall, the Government found a lack of consensus among stakeholders for most identified issues, though there was a significant interest in improving transparency for copyright-protected content used in AI development.

The Government conducted its Consultation on Copyright in the Age of Generative Artificial Intelligence (the Consultation) between October 2023 and January 2024 (as previously reported in the E-TIPS® Newsletter here). It received input from a variety of sources, including individual creators and stakeholders across different industries. Notably, the Government received high engagement from stakeholders within the cultural industries and less engagement from those in the technology and AI sectors. 

The Consultation primarily focused on the following three issues:

1. Use of copyright-protected works in the training of AI systems, notably for text and data mining (TDM) activities. Stakeholders were divided on this point, with creators emphasizing that AI training on copyrighted works without consent or compensation violates their rights, and user groups (including members of the technology industry) taking the position that TDM activities simply involve machine-learning facts and does not actually engage copyright law. Many stakeholders also called for greater transparency requirements to disclose the data being used for AI training.  
2. Authorship and ownership rights related to AI-generated content. Generally, stakeholders indicated that only AI-generated content with sufficient human contribution should be protected under copyright law.
3. Questions of liability, notably when AI-generated content infringes copyright. Stakeholders were divided on whether the existing copyright infringement framework is appropriate for AI-generated content. Some stakeholders wanted clearer liability rules to allocate responsibility in AI copyright infringement matters, while others took the position that the existing copyright framework was sufficient. However, most stakeholders sided with implementing some type of transparency requirement for inputs used to train AI, which would assist rights holders to better exercise copyright remedies, where applicable.

For the full report, click here.

Summary By: Imtiaz Karamat