Canadian News Publishers Sue Anthropic Over Copyright Infringement A coalition of Canada's largest newspapers files an Ontario lawsuit against AI firm Anthropic, accusing the company of copying and using their copyrighted articles without license to train its Claude language model. The case underscores tensions between traditional media and AI developers amid calls for clearer copyright regulations. OTTAWA-A coalition of major Canadian news publishers, including Postmedia, Torstar, The Globe and Mail and The Canadian Press, has taken legal action against the artificial‑intelligence firm Anthropic in an Ontario court. The group alleges that Anthropic copied vast amounts of their copyrighted articles to train its new large‑language model named Claude, thereby infringing on intellectual property rights and enriching itself at the publishers' expense. The filing, dated October 6 and first reported by Cartt.ca, claims that Anthropic's model was built without any valid licence from the plaintiffs, which provide a substantial share of Canada's daily news output and serve millions of readers across print and digital platforms.Anthropic Accused of Systematic Data ScrapingThe complaint details that Anthropic breached the terms of service of the news websites and bypassed security measures designed to prevent copying. It further alleges that the company engaged in "scraping" - systematically accessing and reproducing large volumes of text from the plaintiffs' sites - for the purpose of ingesting data into its AI services.The plaintiffs assert that this practice was repeated on an ongoing basis in response to user requests, and that Anthropic has made no effort to stop the behaviour despite criticism from other content creators and additional lawsuits. The claim also notes that the same publishers, together with CBC and Radio‑Canada, previously filed a lawsuit against OpenAI in 2024, which remains unresolved.These filings come as media organisations push the federal government for clearer regulations around the use of copyrighted content by AI systems, following the national AI strategy's failure to address the issue. The publishers argue that Anthropic's commercial success in the AI sector, evidenced by a growing suite of services and significant capital raised, hinges on unlicensed use of their material. The lawsuit seeks damages, an injunction to halt further infringement, and a license to use their content.Publishers Push for Clearer AI RegulationsAlthough the allegations have yet to be proven in court, the case underscores an emerging conflict between traditional media and AI developers, particularly in a landscape where copyrighted text is used to train algorithms without explicit permission or remuneration. The Canadian Press first reported these developments on October 16, 2026.If the court rules in favour of the publishers, it could set a precedent for how AI firms must comply with copyright law in Canada, potentially requiring licensing agreements or royalty payments for the content used to feed language models. The outcome will also likely influence similar disputes worldwide, as other media houses face comparable challenges from AI developers who rely on large volumes of unlicensed text.The parties are still awaiting a response from Anthropic, which has not yet addressed the claims in court or in the media. The case highlights the broader debate over the responsibilities of AI developers for respecting intellectual property, as machine‑learning systems continue to grow more sophisticated and valuable. Media stakeholders argue that without proper safeguards, their work may be appropriated without compensation, undermining the viability of journalism.Equally, companies like Anthropic contend that training on diverse sources is essential to improving language models' accuracy and usefulness. The legal battle in Ottawa thus represents a pivotal moment in determining how the balance between innovation and copyright protection will shape the future of both industries.