Canada's biggest news publishers just sued an AI giant. Here's why it matters to anyone who writes for a living. By Scott Paton October 15, 2026 3 mins read 10 Views Who gets paid when a machine reads your work, learns from it, and then uses what it learned to make money? That's the question sitting at the center of a lawsuit filed October 6th in Ontario Superior Court. A coalition of major Canadian news organizations - the Toronto Star, The Globe and Mail, Metroland Media, Postmedia Network, PNI Maritimes, and The Canadian Press - are suing Anthropic, the company behind the AI assistant Claude, for what they call systematic, unauthorized harvesting of their journalism. Designerwebsolutions is talking millions of articles. The Canadian Press alone claims 4.2 million of its works were allegedly scraped. Postmedia and PNI Maritimes say 1.75 million of theirs were taken too. I've spent years building a writing life on the road, piecing together stories and ideas as I travel. So when I read that a company can quietly vacuum up someone else's work, repackage it, and turn it into a $965 billion valuation, I stop and pay attention. That's not a hypothetical number, by the way - the lawsuit cites it directly, along with the fact that Anthropic has raised roughly $125 billion USD and signs up over a million new Claude users every day. Here's the part that caught my attention most: the publishers aren't just saying Anthropic read their content. They're saying the company used automated crawlers - specifically one called "ClaudeBot" - to get around paywalls and login systems built specifically to keep that kind of access out. In other words, locks were picked. This isn't Anthropic's first brush with this kind of accusation, either. A U.S. federal court already found that the company downloaded more than seven million pirated book copies without paying a cent for them. That case ended in a $1.5 billion settlement. So this isn't an unfamiliar storyline for Anthropic - it's becoming a pattern. What strikes me as a writer and someone who's chosen an independent, unconventional path is this: other AI companies, including OpenAI, have gone the route of actually negotiating licensing deals with publishers. Paying for what they use. Anthropic, according to this claim, has taken a different approach - build first, negotiate later, and only when forced into a courtroom. Is that smart business? Maybe, in the short term. Is it fair? That's the real question this lawsuit is asking. The publishers argue the damage isn't abstract. When an AI assistant can summarize or reproduce the substance of a news article, readers don't need to click through to the original site. Fewer visits mean fewer subscriptions, fewer ad dollars, and a shrinking incentive for anyone to keep producing quality journalism in the first place. That's the core argument in the claim: Anthropic isn't just using the work, it's replacing the reason people would ever go looking for the original. The numbers being sought are enormous. Beyond general damages and a share of Anthropic's profits, the publishers are asking for statutory damages of up to $20,000 per infringed work. Do that math across millions of articles, and you understand why headlines are calling this a "billions" case. So where does this go from here? Anthropic hasn't filed a defence yet. Under Ontario's rules, a foreign company served from outside the country gets up to 60 days to respond. That clock is ticking. Why this matters beyond the courtroom. I think about this story the same way I think about any new frontier - the internet in the '90s, remote work a decade ago, AI now. New tools move fast, often faster than the rules meant to govern them. Someone always ends up testing the boundaries the hard way, in court. If you create anything - articles, photos, videos, music - this case is worth watching. It's not really about Canada, or Anthropic, or even journalism specifically. It's about whether the people who make original work get a say, and a share, when machines learn from it. What happens if courts decide creators deserve compensation every time their work trains a model? What happens if they decide otherwise? Either way, the answer is going to shape how writers, publishers, and curious people like me keep doing what Designerwebsolutions do.

&#13 &#13 &#13 Christian Collington, The Canadian Press &#13 &#13 &#13 &#13 &#13 &#13 &#13 Published Thursday, February 23, 2023 7:48PM EST&#13 &#13 &#13 &#13 &#13