OTTAWA — The government should require that content generated by artificial intelligence be clearly labelled, a House of Commons committee said in a new report.
The members of Parliament on the committee are calling for standardized labels for AI content that are visible and that the public can understand. They say the requirement should apply to all relevant sectors, including digital platforms and broadcasters.
This would “promote transparency, maintain public trust and preserve the integrity of Canada’s information and cultural ecosystem,” the report said.
It called on the government to establish “a framework governing the systematic and easily identifiable labelling of content created with the assistance of artificial intelligence, including through the use of metadata, digital watermarks or other robust technical solutions.”
AI labelling is among 13 recommendations in a report released this week by the committee, following a study of the impacts of artificial intelligence on the cultural sector.
Most of the testimony the committee heard expressed “profound concerns” about how AI could threaten the long-term viability of the creative sector, the report said.
It cited witness testimony about the use of copyrighted works to train AI models without authorization or compensation, “the proliferation of synthetic content on distribution platforms,” the way AI is competing with creative professionals, “the elimination of jobs and even entire occupational categories” and “the erosion of cultural sovereignty and cultural diversity under the dominance of tech giants.”
The committee said the government should ensure the Copyright Act applies to AI-generated content and require that AI developers be more transparent about the use of copyrighted works to train their models, to allow for “proper authorization and licensing.”
It also called on the government to “establish a clear opt-in consent requirement for the use of copyrighted works in the training of artificial intelligence systems, ensuring that creators’ works may not be used for text and data mining or model development without their prior authorization.”
This report by The Canadian Press was first published April 16, 2026.
Anja Karadeglija, The Canadian Press
