Canada Signals Tougher Stance on AI Use of News Content Amid Growing Industry Concerns

Canada’s federal government is raising fresh concerns over how artificial intelligence (AI) systems use news content, with Culture Minister Marc Miller calling for urgent discussions on accountability and fair compensation.

Speaking at a national summit focused on AI and culture, Miller warned that the way AI platforms currently process and reproduce news risks undermining the very purpose of journalism. He described the practice as one where original reporting is “reused and reshaped,” potentially weakening its value and intent.

His remarks come amid increasing scrutiny of AI companies and their relationship with news organizations. While Canada’s Online News Act already requires major tech platforms to compensate publishers for displaying their content, the question now is whether similar expectations should apply to AI developers.

Miller clarified that the issue is not necessarily about expanding existing laws, but about ensuring responsible behavior across the industry. He emphasized that the core principle remains unchanged: creators should be fairly compensated when their work is used.

The debate has gained momentum following a recent study from researchers at McGill University’s Centre for Media, Technology and Democracy. The study examined how leading AI tools including ChatGPT, Gemini, Claude, and Grok handle Canadian news content. After testing more than 2,200 news articles, researchers found that AI systems failed to properly cite sources in approximately 82 percent of cases.

According to the report, AI platforms are extracting value from journalism at multiple stages. This includes using archived articles to train models, generating summaries or derivative content without attribution, and providing direct answers to users that may reduce the need to visit original news websites.

Researchers argue that this trend could further weaken the financial stability of journalism, an industry already struggling with declining revenues. Unlike social media platforms, which primarily redirected audience attention, AI tools are now delivering news content directly, often eliminating the need for readers to engage with the original source altogether.

Meanwhile, tensions between media organizations and AI companies are escalating in the courts. A coalition of major Canadian news outlets including The Canadian Press, Torstar, The Globe and Mail, Postmedia, and CBC/Radio-Canada has filed a lawsuit against OpenAI. The publishers allege that their copyrighted material has been used without permission to train AI systems, generating profits without compensation.

The legal battle, launched in late 2024, could take years to resolve and may set a significant precedent for how copyright laws apply to AI technologies.

On the issue of copyright, Miller acknowledged its complexity and cautioned against rushing into legislative changes. He noted that intellectual property reform is a long-term process but stressed that existing laws should still protect creators and ensure they are properly compensated.

AI companies, however, have argued in past consultations that using publicly available material for training purposes does not constitute a violation of copyright law.

As the conversation continues, pressure is mounting from creative industries, unions, and media organizations to introduce clearer rules potentially including licensing systems to regulate how AI systems use copyrighted content.

For now, the government appears focused on encouraging responsible practices rather than immediate legislative expansion, while signaling that the broader debate over AI and intellectual property is far from settled.