A new report commissioned by the Australian government finds that enforcing a social media ban for children under 16 is technically possible but carries significant risks, including privacy concerns and accuracy issues with age verification technologies. The study, conducted by the UK-based Age Check Certification Scheme, was published on Sunday and highlights the challenges ahead of the ban’s implementation in December.
The report tested 60 technologies from 48 vendors, including methods like ID verification, facial age estimation, parental approval, and biometric analyses. It concluded that no single solution is universally effective, and all methods have shortcomings, necessitating a layered approach for robustness. For instance, ID-based verification is the most accurate but raises privacy risks due to potential data retention and sharing with regulators, which could lead to breaches in a country already scarred by high-profile data incidents.
Facial age estimation technology, used by platforms like Meta and Snapchat, showed 92% accuracy for adults but has a “buffer zone” of two to three years around age 16 where errors increase. This results in false positives (incorrectly allowing underage users) and false negatives (blocking eligible users), with false rejection rates of 8.5% for 16-year-olds and 2.6% for 17-year-olds. The report emphasized that errors are inevitable and fallback options, such as additional verification steps, will be needed.
Privacy concerns are a major focus, with evidence that some technology providers might over-anticipate regulatory needs, building tools that could compromise user data. The report warned that this could exacerbate risks in a landscape where data breaches have become common. Additionally, biases were identified in age estimation systems, with lower accuracy for non-Caucasian users, females, and Indigenous people due to underrepresented training data, highlighting equity issues.
Communications Minister Anika Wells responded by stating that there is no “one-size-fits-all” solution but insisted that social media companies, with their advanced AI and resources, must implement effective age checks by the December deadline. Under the new laws, platforms face fines of up to A$50 million if they fail to take “reasonable steps” to prevent under-16s from having accounts, though these steps are yet to be fully defined.
The ban, touted as a world-first, is popular with many Australian parents who support measures to protect children from online harms like cyberbullying and exposure to harmful content. However, mental health advocates and experts criticize it, arguing that it could isolate children, cut them off from positive connections, and push them towards less-regulated parts of the internet. They suggest alternative focuses, such as better content moderation and digital literacy education.
The report also addressed circumvention tactics, noting that vendors are developing tools to combat VPN usage and AI-generated forgeries, but these approaches have their own challenges related to accuracy and privacy. As Australia moves forward, the findings will likely influence global debates on youth online safety, emphasizing the need for balanced solutions that safeguard both protection and privacy.