A government-commissioned report has raised red flags over the reliability of selfie-based age estimation software, which is set to underpin Australia’s new ban on under-16s using social media from December.
The report, released Monday, found the technology was generally accurate, fast and privacy-preserving, but flagged “unacceptable” error rates for certain groups — particularly non-Caucasian users, teenage girls, and those close to the minimum age threshold of 16.
While the system performed strongly for adults over 19, the study found a “grey zone” within three years of the age cut-off where results became far less certain.
Sixteen-year-olds faced an 8.5% risk of being wrongly classified as underage and could be pushed into additional verification methods such as ID checks or parental consent.
The findings raise concerns for platforms including Instagram, YouTube and TikTok, which under the new law must prove they are taking “reasonable steps” to block underage users or risk fines of up to A$49.5 million (US$32 million).
The report suggests some users may experience smooth onboarding, while others could face repeated verification hurdles — foreshadowing an uneven rollout when the ban takes effect later this year.
“While systems generally performed well across diverse user groups, some showed reduced accuracy for older adults, non-Caucasian users and female-presenting individuals near policy thresholds,” the report added.
Social media experts warned the findings cast doubt on whether the under-16 ban can be effectively enforced within just three months, raising concerns about the government’s rollout timeline and the readiness of the technology.
According to Justine Humphry, a media researcher at the University of Sydney who specialises in online safety there are a lot of variations in accuracy.
“That variation is concerning, with the tight schedule for the introduction of a system that will need to be robust and working by the end of this year,” says Humphry
Communications minister Anika Wells says the report proved that “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”.
As the December deadline approaches, the government faces mounting pressure to balance child safety with technological limits, leaving platforms and parents bracing for a rollout that may prove far from seamless.
