As we enter the age of AI and its unwavering impact on content creation, search and discovery, Kiara Ferreira and Venelize de Lange from media update explore an economic problem posing as a technical controversy.
Google has raised a recent stance that companies should be allowed to train AI on online content without compensation — and this opinion is placing the company in disagreement with media organisations worldwide.
Google's argument states that the process of training AI models on non-paywalled internet content involves learning patterns, and denies the claim that this includes the reproduction of original work. Therefore, analysing publicly available text to improve AI systems is, according to Google, fundamentally different from republishing original journalism.
Yet, publishers understandably view the situation in a different light. They question what Google's claim of AI producing "new content" truly means when so much of its output simply repurposes existing reporting or research — a grey area that challenges traditional notions of ownership and attribution.
Publishers' concerns are not only about ownership, but sustainability. AI generated summaries appear at the top of search results, offering users instant answers that often remove the need to click through to the original source.
For newsrooms that rely on traffic and advertising revenue, fewer clicks translate directly into shrinking income and reduced capacity to fund journalism.
Additionally, while Google frames publicly accessible content as fair game for training, does that mean that free access translates into free for monetisation? Google's stance demonstrates that content created for human readers automatically becomes a resource for profit-generating AI systems without compensation.
While Google points to opt-out tools that allow publishers to limit how their content is used for AI training, publishers argue that these controls are neither comprehensive, nor risk-free. Opting out may preserve content from AI systems, but it can also threaten visibility in search, a trade-off many publishers simply cannot afford to make.
The debate is now extending beyond commercial disagreement into regulatory territory. Globally, the debate is increasingly drawing the attention of policymakers as questions grow around how AI systems use public content and what protections should apply to publishers.
In South Africa, where AI regulation is still developing, the issue raises urgent questions about how local media organisations will be protected as automated systems mediate access to news and information.
This, however, is not just a legal dispute — it is a reckoning over digital value. Journalistic research and creative labour continue to underpin the online world, even as AI systems reshape how that content is consumed.
If companies like Google can build powerful new products on the back of freely accessible content, the question inevitably presents itself: should "public" still mean "unpaid", or do the rules on which this synonymy is based need urgent revision?
Did you enjoy this article? Let us know in the comments section below.
*Image courtesy of Canva
**Information sourced from Press Gazette, Android Headlines, Digiday, Fluxmans, and The University of Alabama