Several opinion articles report an experiment in which the author poses as a 17-year-old girl online to see what content a recommendation algorithm serves. Instead of finding primarily body-image or eating-disorder material, the authors say the algorithm repeatedly surfaces gender-divisive content, emphasizing division between men and women.
The pieces describe the result as a pattern rather than a one-off recommendation. Across outlets, the core claim is that algorithmic feeds do not simply steer toward one category of harmful content, such as disordered eating, but also promote narratives that separate or polarize genders. The articles focus on the authors’ observations of what appears in the feed and how the content framing contrasts with expectations.
While each source is written from a similar perspective, they differ mainly in the framing of the issue: the Age, the Sydney Morning Herald, and the Brisbane Times present the same central conclusion that recommendation systems can amplify gender-based division, even when the initial intent is to study body-image related content.
