Australia’s Social Media Age Verification Law is Already Failing

A law meant to lock people under 16 years of age is already failing as people are incorrectly passing facial recognition checks.

Australia’s efforts to ban anyone under 16 got off to a predictably rocky start. Billed as a world first, Australia took the especially extreme measure of using age verification to ban anyone under the age of 16 from social media entirely. It’s not a case of trying to prevent younger people from viewing porn or accessing harmful content, it’s just banning them entirely. This under the cult-like belief that no amount of social media is healthy to younger people.

The movement to stop younger people from forming communities online started at least as far back as 2024. The social experiment faced considerable pushback ranging from concerns that the technology is broken to the considerable harm this would have on younger people. Unfortunately, the Australian government took the approach that so many other governments pushing age verification took: dismiss all logic and reason and reject all the evidence and push forward anyway. After all, they doubled down and implemented a $50 million fine for platforms that refuse to go along with this ridiculous idea.

Ironically, a study out of Australia found that age verification was deeply problematic from a security and privacy standpoint. As if to confirm these findings, Discord suffered from a breach, exposing people’s IDs to unauthorized third parties.

Additional research coming out of Australia also confirmed, yet again, that the accuracy of age verification through facial recognition and AI is comically bad. Of course, this is a policy based on personal belief rather than evidence based policy making, so the Australian government spun that research as proof that they have the right idea even though the research in question did no such thing.

Recently, additional research confirmed that facial recognition technology is simply too inaccurate to be trusted. Again, though, the government doesn’t believe in pesky little things like evidence and facts. They got a good feeling about it and, damn it, they are going to run with that.

… and run with it they did. Reports are highlighting the fact that the Australian social media ban law has taken effect. From Al Jazeera:

Australia has banned children under 16 from social media in a world-first, as other countries consider similar age-based measures amid rising concerns over its effects on children’s health and safety.

Under the new law, which came into effect at midnight local time on Wednesday (13:00 GMT on Tuesday), 10 of the biggest platforms face $33m in fines if they fail to purge Australia-based users younger than 16.

The law has been criticised by major technology companies and free speech campaigners, but praised by parents and child advocates.

The Australian government says unprecedented measures are needed to protect children from “predatory algorithms” filling phone screens with bullying, sex and violence.

“Too often, social media isn’t social at all,” Prime Minister Anthony Albanese said in advance of the ban.

“Instead, it’s used as a weapon for bullies, a platform for peer pressure, a driver of anxiety, a vehicle for scammers and, worst of all, a tool for online predators.”

Yup, no amount of evidence is going to put a stop to that personal belief. So, it is probably not a surprise at all that the age verification technology that lawmakers and certain interest groups religiously believe works no matter what the evidence says is already failing pretty spectacularly. From The Guardian:

A 15-year-old boy has been able to pass Snapchat’s visual age check, with the platform warning about the difficulty in enforcing Australia’s social media ban for those aged under 16.

Charlie*, a Sydney-based teen, received a notification last week advising him he needed to go through an age check to keep his Snapchat account as a result of the ban, which takes full effect on Wednesday.

Like many of his peers, the teen said the date of birth on his account – created years earlier – would put him over the age of 16, but he was asked to go through a check regardless.

He chose a facial age estimation technology, provided by tech company k-ID, in which users take a selfie that is then assessed to estimate their age.

Charlie quickly passed as over 16.

“Basically [it said]: ‘Thank you, we won’t bother you any more’,” he said. “I was a bit surprised I got through, to be honest.”

Contacted about the case, a spokesperson for Snapchat said it had previously “expressed concerns about the technical challenges that the government and companies would face in trying to effectively prevent young people from accessing online platforms”.

Oops! It seems that all that annoying evidence may have been right all along. Who would’ve thought?

Unsurprisingly, the platform in question is trying to deflect this whole situation by saying that it shouldn’t be on them to implement age verification in the first place. The critically flawed technology should be foisted onto devices instead of them. It’s not really surprising given that platforms already have a long history of trying to throw the open internet under the bus by pushing to implement age verification at the device level instead. This in a bid to solidify their market dominance at the expense of everyone else.

Obviously, moving age verification over to devices is going to do nothing to make age verification technology any less broken. It just means that the government will have tighter surveillance powers over anyone wanting to use the internet. That, of course, is a further human rights violation (not that governments these days really gives a rats ass about human rights these days, but it’s still worth pointing out nevertheless).

Either way, this is a predictably problem that the Australian government is running into. The government was warned that the technology was broken, but the government ignored all of that and implemented the laws anyway mandating such broken technology. Now, the broken technology is doing what broken technology does – completely and utterly fail to meet the moment. There’s little doubt that age verification apologists will dismiss this latest scandal as little more than the platforms not implementing their broken technology correctly. After all, they used that very excuse to explain away the massive Discord breach everyone with half a sense of how technology works was warning about. In their world view, age verification technology is perfect in every way. It’s clearly the companies fault for not implementing their so-called “perfect” technology. Yeah, there’s no reasoning with those people. Now, we see the problems people like us warned about playing out pretty much like clockwork. It’s only mildly surprising that it is happening so quickly.

Drew Wilson on Mastodon, Twitter and Facebook.