AI in the News: Copilot’s Australian Blind Spots, Health Overviews and Grokipedia in Focus

In a landscape where AI is increasingly curating what we read, a Guardian investigation highlights a problem: AI-generated news summaries from Microsoft Copilot largely sideline Australian journalism, privileging US and European outlets. A University of Sydney study found that only about one-fifth of Copilot’s news prompts linked to Australian media, raising concerns about new forms of media desert and the erosion of diverse voices.

The implications go beyond Australia. If AI systems systematically bias sources toward larger markets, independent voices can struggle to reach readers, and local reporting—already under pressure—could fade from view in everyday AI feeds. The Guardian’s reporting frames this as a potentially dangerous shift in how news is surfaced in the age of generative summaries.

Health information is another frontier where AI \”summaries\” are being used widely. Google’s AI Overviews, introduced as a top-of-page information summary in search results, are said to pull content from a mix of sources. A German study finds that YouTube features heavily in these Overviews for health queries—more than any medical site—sparking questions about accuracy and reliability. The tool is described as providing a quick, confident answer, with mentions of reputable entities like the Centers for Disease Control and Prevention and the Mayo Clinic being cited in some cases. Yet critics warn that the overviews can present a false sense of certainty when the underlying sourcing is uneven or outdated.

In parallel, a Guardian test raised alarms about the latest ChatGPT model citing Grokipedia as a source on sensitive topics, including Iranian political topics and Holocaust-denial material. In a sample, GPT-5.2 cited Grokipedia nine times across more than a dozen questions, prompting concerns about misinformation, bias and the echo chamber risk that comes with AI-selected references. This is a reminder that even highly capable models can replicate misleading narratives if not carefully moderated.

Taken together, these stories underscore a common theme: AI tools can enhance access to information, but they also magnify the responsibility of platforms, researchers and publishers to defend accuracy, transparency and diversity. Readers are urged to approach AI-generated summaries with a healthy level of skepticism, cross-check critical claims, and seek primary sources when possible. The goal is to harness AI as a helpful assistant rather than a sole authority—an approach that keeps public discourse robust as the technology evolves.

As the AI news cycle accelerates, editors, researchers and readers alike must demand clearer signals about provenance, editorial oversight and the inclusion of diverse voices. The daily challenge is not just what AI outputs, but how it decides whom to cite and how to present information so that readers can judge reliability for themselves. This is a story that will evolve week by week, and it helps to stay informed with sources that you trust as you navigate the shifting terrain of AI-driven information.