A trial is testing ways to enforce Australia’s under-16s social media ban. But the tech is flawed

Australia’s move to ban under-16s from social media is receiving widespread praise^[1]. Other countries, including the United Kingdom, Ireland, Singapore and Japan, are also now reportedly considering similar moves^[2].

The ban was legislated^[3] in November 2024 and is due to take effect in December 2025. The law says social media platforms can’t use official IDs such as passports to check Australian users’ ages, and shouldn’t track Australians. But it doesn’t specify the alternative.

To test alternative methods, the federal government commissioned^[4] a trial^[5] of currently available technologies designed to “assure” people’s age online. Run by the Age Check Certification Scheme^[6], a UK-based company specialising in testing and certifying identity verification systems, the trial is in its final stages. Results are expected at the end of June.

So what are the technologies being trialled? Are they likely to work? And how might they – and the social media ban itself – alter the relationship all of us have with our dominant forms of digital communication?

Dead ends for age verification

Age verification confirms a person’s exact age using verified sources such as government-issued IDs. Age assurance is a broader term. It can include estimation techniques such as analysing faces or metadata to determine if users meet age requirements.

In 2023 the federal government rejected mandating^[7] verification technologies for age-gating pornography sites. It found them “immature” with significant limitations^[8]. For example, database checks were costly and credit card verification could be easily worked around by minors.

Nonprofit organisation Digital Rights Watch^[9] also pointed out that such systems were easily bypassed using virtual private networks – or VPNs. These are simple tools that hide a user’s location to make it seem like they are from a different country.

Age assurance technologies bring different problems.

For example, the latest US National Academies of Sciences report^[10] shows that facial recognition systems frequently misidentify children because their facial features are still developing.

Improving these systems would require massive collections of children’s facial images. But international human rights law protects children’s privacy^[11], making such data collection both legally and ethically problematic.

Flawed testing of innovative tech?

The age assurance technology trial^[12] currently includes 53 vendors hoping to win a contract for new innovative solutions.

A range of technology is being trialled. It includes facial recognition offering “selfie-based age checks^[13]” and hand movement recognition^[14] technologies that claim to calculate age ranges. It also includes bespoke block chains^[15] to store sensitive data on.

There are internal tensions^[16] about the trial’s design choices. These tensions centre on a lack of focus on ways to circumvent the technology, privacy implications, and verification of vendors’ efficacy claims.

While testing innovation is good, the majority of companies and startups such as IDVerse, AgeCheck, and Yoti in the trial, will likely not hold clout over the major tech platforms in focus (Meta, Google and Snap).

This divide reveals a fundamental problem: the companies building the checking tools aren’t the ones who must use them in the platforms targeted by the law. When tech giants don’t actively participate in developing solutions, they’re more likely to resist implementing them later.