TLDR: A recent Australian government report has uncovered substantial deficiencies in age-verification technology, particularly selfie-based age estimation software, just months before a nationwide ban on social media for teenagers under 16 is set to commence in December. The study points to accuracy issues, especially around the 16-year-old age threshold, and concerning biases against non-Caucasian users and teenage girls, raising doubts about the practical implementation of the upcoming ban.
CANBERRA, Australia – A government-commissioned report in Australia has cast a shadow over the impending social media ban for teenagers under 16, revealing significant flaws in the age-verification technologies intended to enforce it. The ban, a landmark legislative move aimed at protecting young people online, is slated to begin in December 2025.
The report, released on Monday, September 1, 2025, specifically evaluated selfie-based age estimation software, a primary method considered for verifying users’ ages on major platforms such as Instagram, YouTube, Facebook, X, TikTok, and Snapchat. While the study acknowledged that these systems were “broadly accurate, fast, and privacy-respecting” , it highlighted critical shortcomings that could undermine the effectiveness and fairness of the ban.
One of the most prominent issues identified is a “grey zone” where accuracy sharply declines for users close to the 16-year-old cut-off age . The report found that individuals exactly at 16 years old faced an 8.5% chance of being incorrectly identified as underage . In such instances, supplementary verification methods like ID checks or parental consent would be necessary .
More alarmingly, the study revealed inherent biases within the technology. Non-Caucasian users and teenage girls experienced disproportionately higher error rates, with some groups encountering “unacceptable” levels of inaccuracy . In stark contrast, Caucasian users just above the age threshold faced minimal disruption . Experts noted that age estimation systems commonly exhibit an average error margin of 1.3 to 1.5 years .
Communications Minister Anika Wells acknowledged the challenges but defended the report’s overall findings. “While there’s no one-size-fits-all solution to age assurance, this trial shows there are many effective options and importantly that user privacy can be safeguarded,” she stated .
However, critics argue that the findings underscore the impracticality of the ban. Justine Humphry, a media researcher at the University of Sydney, commented, “It seems like there are a lot of variations in accuracy” . Associate Professor of Law Faith Gordon from the Australian National University asserted that the ban is “not a watertight solution at all” and that age-assurance technology is “clearly not the ‘silver bullet’ to make the digital world safer for children” . She also pointed out the known biases of facial-recognition technology against non-white or non-male presenting individuals . Greens Senator David Shoebridge echoed these concerns, stating, “The age-assurance trial findings accidentally prove the social media age ban is unworkable and it is time to rethink this flawed approach” .
Also Read:
- Study Reveals Significant Gender Disparity in Artificial Intelligence Adoption
- Meta Faces Backlash Over Unauthorized AI Chatbots Impersonating Celebrities
Social media platforms, including those owned by Meta and Alphabet, are now confronted with the formidable task of implementing these safeguards. Failure to comply could result in substantial fines, reaching up to AUD $49.5 million (approximately US$32 million) . The legislation, however, presents a paradox: while document-based age verification is considered the most reliable method, it is explicitly prohibited for platforms to use it . This leaves platforms reliant on less precise age estimation systems, further complicating the rollout of the ban.
