Canadian Heritage

The House of Commons Standing Committee on Canadian Heritage met to study the impact of artificial intelligence on Canada’s creative industries. Witnesses included Stephanie Enders from the Alberta Machine Intelligence Institute (AMII), Pierre-Philippe Côté from Orio, Archita Ghosh from Rezolution Pictures International Inc., Victoria Shen and Neal McDougall from the Writers Guild of Canada, and Tania Kontoyanni from the Union des Artistes.

Stephanie Enders emphasized that AI literacy is a fundamental requirement for all workers, including those in cultural industries, and that Canada has been slower than G7 peers to adopt AI. She recommended targeted investments in research, access to capital and compute for AI-first start-ups, and the use of tools like model cards and data sheets to improve transparency. She argued that AI can support the business of creative industries, not just creative outputs, and that investing in the creative sector is an investment in the national dialogue around technology.

Pierre-Philippe Côté highlighted the potential for AI to stimulate creativity and expand Canada’s cultural reach, but stressed the need for robust legal and ethical frameworks to protect copyright and ensure digital sovereignty. He warned that excessive dependence on American technologies, citing laws like the Patriot Act and CLOUD Act, risks Canadian cultural sovereignty, and recommended supporting local, decentralized, and sustainable cloud infrastructure like Orio. He disagreed with the notion that artists seek handouts, arguing they are entrepreneurs who contribute to society and innovation.

Archita Ghosh described how Rezolution Pictures uses AI for productivity but raised concerns about indigenous data sovereignty, including the scarcity of accurate cultural datasets and lack of transparency in AI training sources. She recommended creating smaller, vetted, and curated datasets with consent, credit, and compensation for indigenous creators, and noted that stereotypes and biases in AI reflect unsupervised online data. She did not directly disagree with other witnesses but emphasized the unique challenges of indigenous representation.

Victoria Shen and Neal McDougall argued that generative AI threatens to flood the media ecosystem with homogenized content and undermine human creators. They recommended that copyright law require consent, credit, and compensation for AI training inputs on an opt-in basis, with no fair-dealing exception or compulsory licensing, and that AI outputs remain uncopyrightable. They also urged that cultural funds like the Canada Media Fund and Telefilm Canada should not support AI-generated content, and that transparency requirements be imposed on AI companies.

Tania Kontoyanni stated that artists are threatened with extinction by AI, as their works, voices, and images are appropriated without consent. She called for absolute transparency, mandatory labelling of AI-generated content, and the right for artists to consent to use of their works and image. She recommended modernizing the Copyright Act, updating private copying rules, and providing support for artists to defend their rights against tech companies. She disagreed with the idea that AI offers positive benefits for performers, arguing that their identity is at stake.

During the question period, members explored topics including AI literacy, data sovereignty, the threat to dubbing and voice work, and the need to protect indigenous cultural data. The committee also debated scheduling and a motion by Martin Champoux to study the effects of online content on children and adolescents, which was adopted with an amendment for a government response. The committee agreed to hold a business meeting on November 17 to discuss a long-term schedule and provide instructions for the report on the AI study.