AI and Journalism: A One-Way Street
A new report has spotlighted a growing tension at the heart of Canada's media landscape: artificial intelligence systems are routinely drawing on Canadian journalism to generate responses and summaries, but they seldom credit, link to, or compensate the news organizations that produced that content.
For an industry already struggling with declining ad revenues, shrinking newsrooms, and audience fragmentation, the findings add a new dimension to a crisis that Canadian journalism has been navigating for years.
What the Report Found
According to the report, AI systems, including large language models and AI-powered search tools, frequently incorporate facts, framing, and analysis that originated with Canadian journalists and media outlets. When users ask these systems questions about Canadian current events, politics, or local news, the responses are often built on journalism that required significant time, expertise, and resources to produce.
Yet in the vast majority of cases, these systems provide no attribution, no link to the original article, no mention of the outlet, no mechanism for the news organization to be compensated for the use of its work.
Why It Matters
The issue goes beyond credit and courtesy. It strikes at the economic foundation of journalism. News organizations rely on readers visiting their websites, subscribing to their publications, and engaging with their advertising to generate revenue. When AI systems can synthesize and deliver journalism without routing users back to the source, that traffic, and the revenue it represents, is lost.
Canada's Policy Response
Canada has been more proactive than many countries in addressing the relationship between news publishers and large tech platforms. The Online News Act, which came into force in 2023, requires major digital platforms to negotiate compensation agreements with Canadian news organizations for the use of their content.
But the law was primarily designed with social media and news aggregation in mind, not the newer wave of generative AI systems that are transforming how people find and consume information. The report's authors are calling for updated policies that specifically address how AI companies use journalistic content.
The Attribution Question
Some AI developers argue that their systems don't "copy" journalism in any meaningful sense. That they learn patterns from text during training rather than storing and reproducing specific articles. Critics counter that this distinction is legally and ethically murky, and that regardless of the technical mechanism, the value flowing to AI companies derives in part from journalistic work.
Several major news organizations internationally have launched legal action against AI companies over these issues, and Canadian publishers are watching those cases closely.
What Comes Next
The report is expected to fuel ongoing discussions in Ottawa about how to modernize Canada's digital media policy for the AI era. Journalism advocacy groups are calling on the federal government to extend the principles of the Online News Act to cover AI systems and to ensure that Canadian newsrooms can negotiate fair terms for the use of their content.
For readers, the message is straightforward: if you value Canadian journalism, seek it out directly, and consider supporting the outlets that produce it.
Source: Google News Canada
