Ottawa’s national artificial intelligence strategy arrives with billions in funding and bold ambitions but for creators, journalists, and publishers, one glaring omission overshadows the rest.
When the Liberal government unveiled its long-awaited national AI strategy earlier this month, the announcement carried the weight of a landmark moment. A $2.3-billion commitment. A stated mission to deepen Canadians’ engagement with artificial intelligence. Fanfare from tech corners across the country.
What it did not carry was the word “copyright” not once in fifty pages.
For the news, music, book, and film industries, that silence was deafening.
“You’ve got basically a 50-page document that came out and the word ‘copyright’ didn’t appear once,” said Paul Deegan, CEO of News Media Canada, whose organization represents publishers from coast to coast. “Which is troubling to news publishers but it’s also troubling to music publishers, book publishers and others.”
The grievance goes beyond semantics. At its core, the concern is about what happens when AI companies hoover up the creative and journalistic output of an entire culture to train their systems without asking, without paying, and without acknowledgment.
Marie-Julie Desrochers, executive director of the Coalition for the Diversity of Cultural Expressions, put it plainly. Creators across sectors film, television, music, books are still watching their work disappear into the training pipelines of major tech firms, with no consent sought and no compensation offered.
“That needs to change because it’s the foundation of our ecosystem,” she said. “If creators don’t get paid for creating, then it’s a huge problem.”
Her coalition represents more than 50 organizations. The breadth of the concern, in other words, is not niche it cuts across virtually every cultural industry in the country.
What both News Media Canada and the Coalition for the Diversity of Cultural Expressions are requesting from Ottawa is not, on its face, complicated. They want the government to declare that it will not introduce a broad exemption to copyright law permitting companies to use protected content for AI training purposes.
Taylor Owen, founding director of the Centre for Media, Technology and Democracy at McGill University, called the absence of any copyright language in the strategy “conspicuous.” His centre published a report in March documenting how AI systems draw heavily on Canadian journalism for the information they deliver to users without offering attribution or compensation in return.
The ask also aligns with a recent recommendation from the House of Commons heritage committee. Following its own study of AI’s impact on the cultural sector, the committee called on the government to require opt-in consent before any copyrighted work could be used for text and data mining or AI model development.
While Ottawa deliberates, the legal system is moving. A coalition of prominent Canadian news organizations among them The Canadian Press, Torstar, The Globe and Mail, Postmedia, and CBC/Radio-Canada has launched a lawsuit against OpenAI in Ontario, alleging unauthorized use of their published content. The case was filed in 2024, though no verdict has been issued.
South of the border, American publisher A.G. Sulzberger of The New York Times characterized the situation in stark terms this month, calling the mass ingestion of journalistic and creative work an “original sin” and a brazen act of intellectual property theft at a scale the world has never seen. The Times has poured over US$20 million into its own legal battle.
A separate U.S. lawsuit names Cohere a Canadian AI company that has received government support as a defendant, with publishers alleging the firm scraped their articles without permission.
This is not new territory for Ottawa. In 2024, the federal government ran an extensive public consultation on AI and copyright. Creators and publishers showed up in force, demanding action. AI companies pushed back, with some, including Google, arguing that training on published content already falls within existing copyright exemptions and even suggesting the government codify that interpretation explicitly.
Florian Martin-Bariteau, research chair in technology and society at the University of Ottawa, said he fully expected the national AI strategy to address what emerged from that consultation process.
“We know that the government has been working on this,” he said. “I think some people were maybe expecting some policy-making, maybe a reform of the Copyright Act.”
Instead, the strategy arrived without a word on the subject.
There are signs the government’s private position may be closer to what creators are asking for than the silence of the strategy suggests. Culture Minister Marc Miller, when pressed in March on why Canada still has no stated position on AI and copyright, appeared to push back against the idea of opening copyright law to accommodate AI companies.
“Current copyright law does and should protect those that have created material and people need to be compensated properly,” Miller said at the time.
He declined, however, to answer a follow-up question about whether Canada would oppose a text and data mining exemption if it emerged as a topic in trade negotiations with the United States. A spokesperson for the minister confirmed this week that there are no updates on the government’s plans.
Deegan and Desrochers both read Miller’s earlier comments as an implicit signal. But they want something more binding a clear public statement that would unlock the next phase of what they believe is necessary: a functioning licensing market where rights holders grant consent and receive payment for how their work is used.
Australia has already made that statement, they noted. Canada has not.
Some experts urge caution about reading Ottawa’s hesitation as indifference. Teresa Scassa, Canada research chair in information law and policy at the University of Ottawa, points to a broader international dynamic. Copyright is woven into a web of global treaties and trade agreements, she explained, and what Canada does unilaterally can carry ripple effects.
“It’s quite possible Canada’s hanging back and waiting to see where the tide goes on this set of issues,” she said, “simply because we’re just not big enough to be charting a new or different path on these copyright questions.”
That calculus might be strategically sensible. But for journalists, authors, musicians, and filmmakers watching their work train the systems that may one day replace them patience is wearing thin.
Deegan also raised a pointed argument about consistency: if Ottawa is actively championing Canadian AI companies like Cohere, it ought to hold them to the same copyright standards it expects everyone else to follow.
“When the government is doing everything, they can to help support national champions,” he said, “I think they should ensure that those companies they are championing are living by Canadian copyright laws.”
The strategy is out. The money has been announced. Now, the industry wants an answer to a question that billions of dollars in funding has so far failed to address: whose work is it, anyway?
