TL;DR: Canadian news publishers sue Anthropic in an Ontario court: Postmedia, Torstar, The Globe and Mail and The Canadian Press say the Claude maker infringed their copyright by scraping their journalism to build and run its AI models. The statement of claim was filed on October 6, 2026 and became public on October 19. The publishers also allege Anthropic breached their websites’ terms of service and got around technical protections meant to stop copying. None of the allegations have been proven in court, and Anthropic had not responded to requests for comment when the case was reported. The same group is already suing OpenAI in a case filed in 2024 that is still unresolved.
Canadian news publishers sue Anthropic over how its Claude models were built, and the case adds Canada’s biggest newspaper groups to the list of publishers taking AI companies to court. The lawsuit, first reported by Cartt.ca and then by CBC News and The Globe and Mail (both carrying a Canadian Press report), argues that Anthropic made money from Claude without ever licensing the news content it allegedly copied. Below we cover who is suing, what they claim, what is still unknown, and how the case fits into the wider fight over AI training data.
Who is suing Anthropic, and where
The plaintiffs describe themselves in the claim as the “News Media Companies.” Together they publish a large share of Canada’s daily journalism.
| Plaintiff | What it publishes |
|---|---|
| Postmedia | Canada’s largest newspaper chain, including the National Post and many big-city dailies |
| Torstar | Owner of the Toronto Star and other Ontario titles |
| The Globe and Mail | National daily newspaper based in Toronto |
| The Canadian Press | National news agency that supplies stories to outlets across the country |
The case was filed in an Ontario court on October 6, 2026. According to the Canadian Press report, the claim says these outlets “are responsible for generating a significant portion of Canada’s journalistic content – from daily print broadsheets to round-the-clock digital news coverage,” and that they collectively have millions of readers.
What the publishers allege
The statement of claim makes several separate allegations. Taken together, they go beyond the usual “you trained on our articles” complaint.
1. Copyright infringement through training
The core claim is that Anthropic copied the publishers’ content to develop Claude. The claim says: “To obtain content for ingestion into its models and services, Anthropic deliberately ‘scrapes’ (i.e. accesses and copies) enormous amounts of text data from the internet and from the Plaintiffs’ Websites.”
2. Ongoing scraping in response to user prompts
The publishers also say Anthropic keeps scraping their content on an ongoing basis when users ask Claude questions. That targets live retrieval, not just a one-time training run. If it holds up, the claim could matter for how AI assistants fetch and summarize news in real time.
3. Breach of terms of service and circumvention
The claim alleges Anthropic breached the news sites’ terms of service and circumvented “technological protection measures” the outlets had put in place to restrict copying. Canada’s Copyright Act has specific rules on circumventing such measures, so this part of the case may be argued separately from the infringement claim.
4. Unjust enrichment
The publishers argue 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.” They add that Anthropic “has been substantially and unjustly enriched to the detriment of the News Media Companies.”
The claim also says that despite “numerous other lawsuits and public criticism, Anthropic has done nothing to curtail its unlawful behaviour.” Again, these are the plaintiffs’ allegations. A court has not tested any of them.
What we don’t know yet
- Damages: The published reports do not give a dollar figure for what the publishers are seeking.
- Anthropic’s defence: Anthropic had not responded to a request for comment when the Canadian Press reported the case, and it has not yet filed a defence in public.
- Which content and which models: The reports don’t say which articles, time periods or Claude versions the claim covers.
- Timeline: Cases like this usually take years. The publishers’ 2024 lawsuit against OpenAI is still unresolved.
The same publishers are already suing OpenAI
This isn’t the group’s first AI case. In 2024 the same companies, together with CBC/Radio-Canada, sued OpenAI for allegedly using their news content to train ChatGPT. According to the Canadian Press, that case has yet to be settled. The Anthropic suit follows the same strategy and points to an industry-wide approach: Canadian publishers want licences, not one-off disputes.
Publishers in other countries are doing the same. The New York Times is suing OpenAI, and earlier this month we reported that USA Today’s owner sued OpenAI for a quarter billion dollars. The pressure is changing platform behaviour too. Meta recently removed AI scraping language from its terms after publisher backlash, and Google has been showing more links in AI Overviews after similar complaints.
Anthropic’s copyright history
Anthropic has already paid heavily over training data once. In the US, a federal judge gave final approval in July 2026 to Anthropic’s $1.5 billion settlement with authors in Bartz v. Anthropic, a class action over books used to train Claude, AP reported. That case was about books, not news, and was governed by US law, so it sets no precedent in Ontario. It does show that copyright claims against AI developers can carry real financial weight.
The new suit also lands as Anthropic expands in Canada. Days ago the company signed a 20-year lease at a data centre campus in Nova Scotia, which we covered in Anthropic’s Nova Scotia data centre deal.
Why this case matters
| Issue | Why it’s worth watching |
|---|---|
| Canadian law, not US law | Canada uses “fair dealing,” which is narrower and differently structured than US “fair use.” Arguments that worked in US courts may not carry over. |
| Live retrieval claim | Targeting scraping “in response to user prompts” could affect how AI assistants browse and summarize news. |
| Anti-circumvention | A claim that technical protections were bypassed is separate from infringement and may be easier or harder to prove depending on the evidence. |
| Licensing leverage | Two parallel suits against the two biggest AI labs give Canadian publishers a stronger hand in any licensing talks. |
There is also a policy angle. According to the Canadian Press, news publishers have been pushing Ottawa to take a position on how AI systems use copyrighted content after the federal government’s national AI strategy didn’t address the issue. A court ruling, or a settlement, could end up doing what policymakers haven’t.
What it means for Claude users
Nothing changes for Claude users right now. The lawsuit doesn’t restrict the product, and early-stage litigation rarely leads to quick injunctions. Over the longer term, cases like this could push AI companies toward paid news licences, more visible citations, or tighter limits on what assistants pull from publisher sites. If you use Claude for news research, it’s still smart to click through to the original reporting, which is the very thing publishers say AI tools are eroding.
We will update this story if Anthropic responds or files a defence. For more on how the company is handling user data, see our explainer on Anthropic’s Claude voice data opt-in.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
