The Government has released a report from a trial of age assurance technologies to enforce the ban on social media accounts for under-16s.
It found these technologies are “practical” and “achievable,” but have some “inevitable” issues.
The trial focused on verification (official documents), estimation (facial ID scans), and inference (online behaviours) as ways of determining someone’s age.
The Opposition has criticised the trial’s timing, saying the report has been released at “10 seconds to midnight”.
Social media ban
The Government’s ban on children under 16 using social media passed Parliament in November, following a public campaign to protect young people’s mental health.
Social media companies, including Meta (Facebook and Instagram) and Snap Inc (Snapchat) could face fines of around $50 million if under-16s make an account on their platforms.
YouTube was initially exempt, but has since been included. Messaging apps and online gaming platforms won’t be covered by the ban.
Enforcement
The ban is scheduled to come into effect on 10 December.
Enforcement will largely fall on social media companies, not individual users.
The eSafety Commissioner will need to provide guidance on the “reasonable steps” companies can take to stop under 16s accessing their sites.
The Government commissioned an independent firm, the Age Check Certification Scheme, to review the effectiveness of age assurance technologies.
Verification
The first method of age assurance would involve using official identity documents to verify a user’s age.
This could include passports, birth certificates, and photo IDs.
Driver’s licences could also be used. In most of Australia, the minimum age to receive a learners’ licence is 16.
Some third-party platforms require users to take a photo next to their ID in order to validate its authenticity.
Estimation
Age estimation would require using scanning technology to detect someone’s age via facial, voice, and behavioural analysis.
More than 28,000 facial image tests were conducted in Australia as part of the research.
Your contribution ensures The Daily Aus can continue doing the work you love.
I want to give $10 monthly to help TDA keep doing the news.
Your details:
It found the ‘false rejection rate’ for those just outside the ban range were “above acceptable levels”. For 16-year-olds, it was 8.5%.
2.6% of 17-year-olds were mistaken for being under 16.
Inference
Age inference uses existing online data to determine someone’s likely age.
This could include scanning a person’s digital footprint, such as their email addresses and online behaviours.
It would also use official data to detect whether someone is over 16, such as matching a person’s ID with the electoral roll.
School enrolment data would also help determine if a teenager is covered by the ban.
Issues
The Government has flagged it needs to balance online privacy concerns with enforcing age assurance.
The report warned age ’inference’ technology may lead to an unethical level of data collection about individuals.
It also looked at accuracy issues.
The report said there was a “buffer zone” of about 2-3 years, meaning some 16 year olds could be misidentified as being 13, and vice versa.
The report noted “skin tone” could impact the tech’s effectiveness, finding “reduced accuracy for… non-Caucasian users”.
“Underrepresentation of Indigenous populations in training data remains a challenge,” the report stated.
It also looked at some of the ways kids could dodge age assurance, such as using VPNs (Virtual Private Networks).
VPNs change a user’s IP address to make it appear as though they’re in a different location. It said geolocation technology could circumvent VPNs by comparing someone’s IP address to their regional patterns.
Response
The Government acknowledged there would not be a “one-size-fits-all” approach to rolling out age assurance.
Communications Minister Anika Wells said the report was evidence that “digital platforms have access to technology to better protect young people from inappropriate content and harm.”
Melissa McIntosh, the Opposition’s communications spokesperson, said the report has come at “10 seconds to midnight”. The ban is due to come into effect on 10 December.
