The organisation behind the age assurance technology trial that will inform how to keep under-16s off social media has frozen out key stakeholders amid media leaks and resignations of two members.
The trial’s Iain Corby has also downplayed reporting about inaccuracies with the facial estimation technology – one of the technologies tested in the trial – arguing that it can still be used even if it is out by seven years.
The $6.5m age assurance technology trial, run by the UK-based Age Check Certification Scheme (ACCS), tested various types of technology that could be used by social media platforms and adult websites to keep out under-16s or under-18s, respectively, when Australia’s under-16s social media ban comes into force in December.
The project provided its final report to the communications minister, Anika Wells, at the start of August. The final report is expected to run to 10 volumes and 2,500 pages.
However, the stakeholder advisory board for the trial – which comprises tech companies, child safety advocates, academics and privacy advocates – may not see the final report until Wells releases it in the coming weeks. Two sources close to the board told Guardian Australia the board was not expected to be provided with a copy before then due to leak concerns.
When the trial was announced, the project plan noted that transparency was key to ensuring public trust in the project.
Sign up: AU Breaking News email
“The programme needs to be completed with transparency and ensuring the credibility and confidence of participants, the commissioning department and the Australian public,” the plan stated.
Initially, detailed minutes for the stakeholder meetings were published online outlining disagreements and concerns raised by those involved. But the last several meeting minutes have not been posted, and that is expected to continue for the final meeting as the report is released. The group behind the trial did not publicise that the report had been handed to government, apart from a blog post on the trial’s website.
Two members of the advisory board have also quit. Guardian Australia confirmed that the Electronic Frontiers Australia chair, John Pane, resigned from the board last week, following another resignation reported earlier by Crikey.
In a statement issued last week, Pane criticised the preliminary report findings from June – where the project team claimed that age assurance technology could be “private, robust and effective” – as “strong on hype and rhetoric, and difficult to reconcile with the evidence”.
“These political talking points seem to be a case of ‘selling the sizzle and not the steak’– or perhaps even ‘privacy washing’,” Pane said.
He argued that assessment of the privacy practices of some vendors amounted to checking if the vendor had a privacy policy, and was a “tick-box compliance” exercise.
Pane also said the trial organisers had not confirmed whether the vendors who participated in the trial will permanently de-identify all personal data collected from test subjects.
Tim Levy, the managing director of children’s safety technology company Qoria, resigned from the trial earlier this year. Levy said the voracity of the conclusions of the interim report were “not going to match community expectations and I believe my team of 600 dedicated cybersafety professionals would not like us to be associated with such an unsafe report”.
ACCS’s chief executive, Tony Allen, said Pane’s contribution and the work of the board was welcomed but he said all of the points raised “have been addressed in the full report – and in some considerable depth”.
“It is partly because of those points that it is taking some time to prepare the report (and perhaps more so, the supporting materials) for publication.”
Allen said all data has been anonymised and the personally identifiable information deleted.
He said the trial continues to engage with the board and is working for the next meeting. Allen said a gap in publication of meeting minutes was “no conspiracy” but a “factor of preparation for publication, which takes some time”.
Corby, who is responsible for stakeholder engagement for the trial and is also the executive director of the Age Verification Providers Association, told a podcast published on the industry research and consultancy website Biometrics Update last month that people “need to be patient and get the full 10 volumes [of the report] in the public domain, and then it will be a lot clearer what the trial has found”.
He said the report will be a “bible of data” that will be “quoted around the world”.
“[The report] is albeit done by Age Check Certification Scheme, who are known in the sector, but with very close scrutiny from an advisory board and ethics panel, the government themselves in Australia, [and] Prof Toby Walsh providing independent review of the evaluation approach,” Corby said.
“So it’s been done with a lot of discipline around its independence and validity.”
In June, the ABC reported that tests of facial age estimation technology had estimated a 16-year-old being as old as 37. Corby dismissed this report, stating that errors in age estimation don’t undermine the whole project of age assurance.
“You’re always going to have what we call a buffer age, and that might be three years or it might need to be five. Or for one provider, it might be three, and for a provider with a poorer quality algorithm, it might be seven in order to achieve the same level of accuracy overall,” he said on the podcast.
“But that doesn’t mean to say you have to give up on the social media minimum age bill in Australia because one particular category of technology doesn’t give you an exact answer.”
Corby told Crikey this week the trial would not comment until the final report is released.
