Canada’s Biggest News Publishers Are Suing Anthropic Over Claude: Inside the Copyright Fight

Four of Canada’s largest news publishers have drawn a legal line. Postmedia, Torstar, The Globe and Mail and The Canadian Press filed a copyright lawsuit against Anthropic, the company behind Claude, in an Ontario court on October 6, 2026. It is the Canadian news industry’s most aggressive legal challenge yet over how AI companies get their training data.

The publishers and what they say happened

In their statement of claim, the publishers call themselves the “News Media Companies,” producers of a significant share of Canada’s journalism, from daily print editions to round the clock digital coverage. Their core accusation is direct: Anthropic copied their content to develop Claude without permission and without a licence.

The claim’s language is unusually blunt. “Anthropic has capitalized on the commercial success of its AI models, building an expansive suite of services, and raised significant capital, all without obtaining a valid licence from any of the News Media Companies,” it says, adding that Anthropic has been “substantially and unjustly enriched” at their expense. Those are the filing’s words, as reported by The Canadian Press. None of the allegations have been proven in court, and Anthropic had not responded to requests for comment.

The story became public on October 19, 2026, when The Canadian Press reported on the filing, first spotted by Cartt.ca. The two week gap is normal: statements of claim are public documents, but somebody has to go looking.

Why the claim goes further than training data

Most AI copyright lawsuits look backward, at training. This one reaches into the present: according to the filing, Anthropic keeps scraping the publishers’ websites on an ongoing basis in response to user prompts, pulling fresh content when people ask Claude questions. If a court accepts that framing, it could affect how AI assistants fetch live news, not just how they were built.

The claim also goes beyond copying. The publishers say Anthropic breached their websites’ terms of service and got around technical protections such as paywalls and login checks, which can carry separate legal weight from copying itself. The filing describes the scraping as “deliberate,” a word choice that signals intent.

Anthropic has not answered publicly and has not yet filed a defence, though defendants in civil cases get time to respond after being served.

What the publishers are asking the court for

The relief sought is broad. According to reports on the filing, the publishers want damages, an accounting of profits from the disputed content, and a handover of those profits. Alternatively, they seek statutory damages of up to $20,000 per infringed work, plus punitive damages and permanent injunctions against further scraping.

Reports citing the claim put The Canadian Press alone at about 4.2 million works involved, the largest share among the plaintiffs. Statutory damages are designed to punish rather than measure actual loss, so mass copying cannot be treated as a rounding error. Multiplied across millions of works, even a fraction of the maximum is an enormous number, which is presumably the point. Whether a court would award anything near that scale is a separate question, and settlements usually land far below the opening ask.

The injunction request may matter most. An order stopping Anthropic from scraping these sites would force licences, technical walls around Canadian news content, or both. The publishers have also been pressing the federal government to act on AI and copyright, and they were unhappy that the national AI strategy announced in June 2026 did not address it, according to industry reporting.

This is not their first AI lawsuit

The same four publishers, together with CBC/Radio-Canada, sued OpenAI in 2024 over ChatGPT. That case is still unresolved, which shows how slowly these disputes move. Filing a second suit while the first is pending signals that the publishers treat litigation as a long term strategy rather than a one shot warning.

The timing is pointed too. Anthropic has been deepening its presence in Canada, including plans for Canadian data infrastructure. Suing a company that is actively investing in your country says the publishers believe Canadian courts, not just public pressure, are where this gets settled.

Around the world, news organizations and authors have been taking AI companies to court over training data, with mixed results so far. Each ruling in one jurisdiction becomes an argument in the next, and a decision from Ontario could influence how AI firms treat Canadian content generally.

Why this case could matter beyond Canada

The practical question is simple: should AI companies pay newsrooms for the journalism their products rely on? The publishers say yes. Anthropic, like its peers, has generally treated training on publicly available material as lawful, though it has not yet said that in this case.

The answer will shape both industries. News organizations are under financial strain, and every article that trains a model is one the model can then summarize for users who never visit the publisher’s site. If courts require licences, AI companies gain a new cost layer and newsrooms a new revenue stream. If courts side with the AI firms, the licensing pressure largely collapses.

Nothing is decided yet. The claim is only the opening move: Anthropic will file a defence, evidence will be gathered, and the 2024 OpenAI case shows these fights can run for years.

Frequently Asked Questions

Who exactly is suing Anthropic in this case?

Postmedia, Torstar, The Globe and Mail, and The Canadian Press, who call themselves the “News Media Companies” in the filing. The same group, plus CBC/Radio-Canada, also sued OpenAI in 2024, and that case is still unresolved.

What does the lawsuit accuse Anthropic of doing?

Copyright infringement: copying the publishers’ content to build Claude without a licence. The claim also alleges ongoing scraping in response to user prompts, breaches of the sites’ terms of service, and circumvention of paywalls and login checks. None of the allegations have been proven in court.

What could happen if the publishers win?

According to reports on the filing, they are seeking damages or statutory damages of up to $20,000 per infringed work, an accounting and handover of profits, punitive damages, and permanent injunctions against further scraping. A win could force Anthropic to license Canadian news content or stop scraping it entirely. The case is at an early stage: Anthropic has not yet filed a defence, and similar lawsuits have taken years to resolve.