Private, effective age verification is possible: Australia age assurance technology trial

“Age assurance can be done in Australia and can be private, robust and effective.”

This is the key finding of the Australia’s Age Assurance Technology Trial (AATT), which has published its preliminary findings – a set of twelve observations based on the first phases of the Trial’s evaluation, which “highlight broad patterns and trends seen across all the technologies under test.” Those technologies include a range of methods of age verification and biometric age estimation, as well as age inference.

“Preliminary findings indicate that there are no significant technological barriers preventing the deployment of effective age assurance systems in Australia,” says Tony Allen, project director of the AATT. “These solutions are technically feasible, can be integrated flexibly into existing services and can support the safety and rights of children online.”

The AATT, which is being conducted on behalf of the Australian government by the Age Check Certification Scheme (ACCS) and testing partner KJR, has released the findings in the interest of transparency. They include a reminder that the big takeaway – yes, age verification and age estimation can be done – is a statement of capability, and not a policy position on whether or how it should be used.

But, again: it can be done – without unneeded friction or a disproportionate privacy impact, as long as appropriate safeguards are in place. The findings emphasize “the readiness of Australian service providers and technology partners to meet age-related access requirements in both online and offline environments.”

Among the 53 organizations participating in the trial are Australian Payments Plus (ConnectID) and Austroads, the association of Australian transportation agencies, which has launched a Digital Trust Service to handle mobile driver’s licenses (mDL).

Those interested in learning more about the kinds of age assurance technologies available can refer to the “2025 Online Biometric Age Assurance Market Report & Buyers Guide – UK Edition” from Biometric Update and Goode Intelligence.

Thriving sector understands the gravity of privacy policy

The other preliminary findings can be summarized as follows. The evaluation did not reveal any substantial technological limitations to deploying age verification or age estimation.

The practice statements provided by providers of systems classified as Technology Ready Level (TRL) 7 or above tended to accurately reflect their tech’s capabilities.

There are a “plethora of approaches” for different use cases, but no “single ubiquitous solution” that would work across the board. Parental control and consent systems “can be done and can be effective” – but tend to get less effective as kids develop.

The age assurance sector is thriving, a key note for those keeping tabs on government-led initiatives that could cannibalize private startups. It also understands what’s at stake: “We found robust understanding of and internal policy decisions regarding, the handling of personal information by trial participants.”

There is room for improvement; for one, the tools could be easier to use. “We found opportunities for technological improvement including improving ease of use for the average person.” Systems were “generally secure and consistent with information security standards,” but “the rapidly evolving threat environment means that these systems – while presently fairly robust – cannot be considered infallible.”

On the topic of standards, the AATT says its standards-based approach, including through the ISO/IEC FDIS 27566 Series, the IEEE 2089.1 and the ISO/IEC 25000 series (the Product Quality Model), was the right one.

No demographic bias – but some firms too keen to share data with law

Two preliminary findings deserve special attention. First, the one concerning bias – a frequent bone of contention in discussions about biometrics. The trial found that “the systems under test performed broadly consistently across demographic groups assessed.” Specifically, “despite an acknowledged deficit in training age analysis systems with data about Indigenous populations, we found no discernible difference in the outcomes for First Nations and Torres Strait Islander peoples and other multi-cultural communities using the age assurance systems.”

“We found some systems performed better than others, but overall variances across race and gender did not deviate by more than the permitted tolerances set out in IEEE 2089.1.”

The second relates to the issues of data collection and retention, privacy and surveillance. The AATT points to “concerning evidence that in the absence of specific guidance, service providers were over-anticipating the eventual needs of regulators about providing personal information for future investigations.”

“Some providers were found to be building tools to enable regulators, law enforcement or coroners to retrace the actions taken by individuals to verify their age which could lead to increased risk of privacy breaches due to unnecessary and disproportionate collection and retention of data.”

This finding is bound to stoke fears about facial recognition and other biometric systems being used for police surveillance and increasingly invasive law enforcement practices.

Nonetheless, the findings can be said to be very positive from a technical standpoint. They are subject to change in the final report, which will include more detailed technical assessments, testing outcomes and vendor-specific performance data. But the AATT has reported that it is on track to deliver the final report by the end of June, meaning there is limited time for extensive changes. Trial participants are being given fair opportunity to respond to the findings about their age verification products, and the report will be done once the period for response has closed.

Once the AATT has delivered the report, it will be up to the Australian government to figure out what to do next. The expectation is that a final version will be published in full sometime this year. In the meantime, organizers say, “while the findings will be shared with government agencies and regulators for context, decisions about whether and how to implement age assurance requirements remain a matter for government and Parliament.”

Article Topics

Age Assurance Technology Trial  |  Age Check Certification Scheme (ACCS)  |  age verification  |  Australia age verification  |  biometric age estimation  |  KJR Testing