Ottawa's federal government has unveiled its long-awaited artificial intelligence strategy — but for Canada's news publishers and cultural organizations, the announcement raised more questions than it answered, particularly around the thorny issue of copyright.
What the Strategy Says
The federal AI strategy lays out a broad vision for how Canada plans to position itself as a global leader in artificial intelligence, emphasizing innovation, economic growth, and responsible development. It touches on transparency, accountability frameworks, and investment in AI research.
What it conspicuously sidesteps, critics say, is a clear answer on one of the most pressing legal questions facing creative industries: can AI companies scrape and train on copyrighted Canadian content without permission or compensation?
Publishers and Creators Want Answers
News organizations and cultural advocacy groups across Canada have been pushing Ottawa for months to clarify the legal landscape. With generative AI tools now capable of summarizing news articles, reproducing song lyrics, and mimicking authors' styles, the stakes couldn't be higher for the people who create that content.
Representatives from the news industry argue that AI companies are essentially free-riding on decades of expensive journalism — using it to train models that then compete directly with the outlets that produced it. Cultural groups echo similar concerns, pointing out that writers, musicians, and visual artists have had their work ingested into AI systems without consent or payment.
The absence of explicit copyright guidance in the strategy is being read by many as a delay tactic, or worse, a quiet nod to the tech sector's preferred outcome: the status quo.
A Pattern of Deferrals
This isn't the first time the copyright question has been punted. Canada's Copyright Act was last substantially updated in 2012, well before generative AI entered the mainstream conversation. The government launched a copyright review in recent years, but concrete legislative action has been slow to materialize.
In the meantime, AI companies have continued training on vast datasets scraped from the open web — including Canadian news sites, government publications, and cultural archives — with little legal risk.
What Comes Next
Advocates are urging the government to introduce amendments to the Copyright Act that would require AI developers to license content used in training data, or at minimum to make opt-out mechanisms legally enforceable. Some are pointing to moves in the European Union, where the AI Act includes provisions around training data transparency, as a model Ottawa could follow.
For now, news organizations and cultural groups say they'll keep pressing the government for clarity. Without it, the economic and creative livelihoods of Canada's information and arts sectors remain in legal limbo — waiting on Ottawa to decide whose side it's on.
Source: The Canadian Press via Google News Ottawa
