OTTAWA – Culture Minister Marc Miller emphasized the necessity for the government to engage in a serious discussion regarding the utilization of news by AI systems. He remarked, “Having the news cannibalized and regurgitated undermines the spirit of the use of that news in the first place and the purpose for which it’s used, and we have to have a serious conversation with the platforms that purport to use it, including AI shops.”
During a national summit on AI and culture, Miller was questioned about the potential extension of the Online News Act to AI companies. This act mandates that platforms like Meta and Google compensate media outlets for displaying their content. In response to the act, Meta removed news from its platforms, while Google has complied by making payments.
Miller clarified that the conversation should not focus on amending the legislation but rather on ensuring that companies act responsibly. He acknowledged a new report indicating that AI systems rely on Canadian journalism for information but fail to provide proper compensation or attribution.
Researchers from McGill University’s Centre for Media, Technology, and Democracy evaluated 2,267 Canadian news stories using platforms like ChatGPT, Gemini, Claude, and Grok. The findings revealed that these AI platforms did not attribute sources about 82 percent of the time when responding to inquiries about Canadian news events.
The report criticized AI companies for profiting from journalism without proper recognition, stating they extract value from journalism at various stages, from using news archives as training data to generating derivative content. This practice, according to researchers, contributes to the decline of the journalism sector that these AI systems depend on.
Miller stated he had reviewed the report and emphasized the need for legislation that ensures fair compensation for the use of proprietary materials. When asked if this principle would extend to AI companies, he affirmed that "the principle of proper compensation for the use of proprietary material doesn’t change."
Moreover, he reiterated the government’s willingness to negotiate terms that would allow for the return of news content to Meta’s platforms. In a policy brief, the McGill researchers differentiated between the challenges posed by social media and those by AI systems, stating that while social media platforms capture advertising revenue by aggregating attention around news, AI companies absorb the essence of journalism and deliver it as their own product, thus significantly diminishing the need for consumers to visit the original source.
A coalition of Canadian news outlets, which includes organizations such as The Canadian Press, Torstar, the Globe and Mail, Postmedia, and CBC/Radio-Canada, has initiated legal action against OpenAI in an Ontario court. They argue that OpenAI has used their news content without permission or compensation to train ChatGPT, thereby violating copyright laws.
When Miller was questioned about the government's stance on the legality of using copyrighted materials for AI training, he noted the complexity of intellectual property reform and emphasized that it encompasses more than just AI issues. He mentioned that it would be premature to assert that no changes would occur, noting that the current copyright law aims to protect creators and ensure that they are compensated adequately.
During a 2024 consultation on copyright and artificial intelligence, AI companies claimed that their use of material for system training does not infringe copyright law. The lawsuit from news publishers was filed in late 2024, and it remains uncertain how long it will take for the court to reach a decision on the matter.
The House of Commons heritage committee had previously heard from various groups and unions representing creative industries, expressing concerns about AI’s use of copyright-protected works without permission. They advocate for the establishment of a licensing system to govern such usage.
