The Coalition for the Diversity of Cultural Expressions (CDCE) emphasizes the paramount importance of placing the protection of creators’ and their business partners’ rights at the heart of any AI and culture strategy. These elements are essential not only for sustaining growth in this vital economic sector but also for fostering sustainable innovation. This protection is essential to sustaining the diversity of voices and stories that enrich our society, while cultivating a thriving and resilient cultural landscape driven by human creativity for generations to come.

Representing a broad spectrum of cultural disciplines, the CDCE presents in this document a set of recommendations reflecting a strong consensus across its membership. We nonetheless highlight the importance of adopting sector-specific approaches to better address the distinct challenges and characteristics of each cultural industry. We also encourage strong involvement from diverse communities, including Indigenous peoples, linguistic minorities, and other cultural and regional groups that reflect the rich diversity of the country.

To develop an effective Canadian AI and culture strategy, it is crucial to ensure balanced representation of cultural sector experts, AI developers, policymakers, and other relevant stakeholders at every stage of the process. This collaborative approach will help create a strategy that not only addresses the unique needs of the cultural sector but also fosters innovation and sustainability.

The section below provides a summary of the CDCE’s recommendations on generative artificial intelligence as they relate to the cultural sector. Several of these recommendations were presented during public consultations held in 2024, including those addressing the Copyright Act in the context of generative AI and others presented as part of Bill C-27, Part 3 on Artificial Intelligence and Data (AIDA). Additional recommendations emerged in October 2024 during the roundtables on the Canadian Cultural Data Strategy, organized by Canadian Heritage in collaboration with Mila.

RECOMMENDATIONS

  1. Legislation and regulation

    • Consent: In Canada, it is currently prohibited to use works and productions protected by copyright without the consent of the rights holders. The CDCE requests that the Copyright Act not be amended to allow technological development companies to use protected works, productions, and performances to train generative AI systems without authorization or compensation (text and data mining).
    • Transparency
      • Training Data: Implement legally binding measures that require full disclosure of the training data used in AI systems, going beyond mere summaries.
      • AI System Operations: Ensure transparency in the way generative AI systems operate and make decisions, providing clear information on their decision-making mechanisms.
      • AI-Generated Content: Guarantee that AI-generated content is clearly labeled, so the public is fully informed about the nature of the content they are consuming.
    • Copyright and human creativity: Products resulting from purely mechanical AI processes, without any original human expression, are not “works” protected by copyright or neighbouring rights and should not be classified as such. The CDCE insists that the Copyright Act should not be amended to grant new protections to generative AI. Additionally, the CDCE requests that performers’ performances remain fully protected under the Copyright Act, even when the content performed is AI-generated.
    • Trade agreements: maintaining global cultural exceptions in international treaties. When an approach favoring exceptions for culture by chapter instead of a global one is adopted, chapters dealing with artificial intelligence must be included.
  1. Market-driven solutions

    • Flexible licensing models: The development of various licensing models, including voluntary collective licensing and direct licensing, must be encouraged.
  1. Fostering economic development by building adequate tools

It is essential that the cultural sector plays a leading role in shaping these tools, and the CDCE is well-positioned to take on a central role in this effort, provided sufficient financial support is granted. These tool – which could be unveiled at major events in 2025, such as Mondiacult 2025 in Barcelona, Spain – have the potential to position Canada as a global leader. Their development needs to begin without delay.

    • AI Cultural Impact Assessment Framework and Study: We propose establishing a formal assessment to evaluate the effects of AI on employment, artistic creation, the economy, and cultural consumption. This framework would involve developing indicators to continuously monitor these impacts, guiding strategies for preservation, adaptation, and innovation, ensuring that AI supports a sustainable and thriving cultural ecosystem. It could also facilitate the implementation of Canada’s first AI impact studies on the cultural sector.
    • Ethical Framework: We call for the development of a comprehensive ethical framework, specifically designed for AI use in the cultural sector, to guides all strategic decisions from both public and private entities; to reinforce the need to respect human rights, including copyright; and to implement all the necessary tools for the enforcement of these rights. This will allow those developing AI systems to do so in a clear legislative environment, both supporting the economic development of the cultural sector and fostering innovation.
  1. Education and Awareness

    • Ongoing AI Training for the Cultural Sector: Continuous training is crucial to ensure that cultural stakeholders keep pace with the rapid advances in AI technology. This will help them stay informed on AI regulations, operational mechanisms, consent processes, and remuneration models, enabling the cultural sector to adapt and thrive in a rapidly evolving technological landscape.

    Insights from the Coalition for the Diversity of Cultural Expressions following the Roundtable Organized by Heritage Canada in partnership with Mila Towards the Development of a Canadian Cultural Data Strategy for AI

    Report
    CDCE
    1 November 2024