Already and this is the Daily This is the Daily OS. Oh, now it makes sense.
Good morning and welcome to the Daily ODS. It's Tuesday, the second of September.
I'm Harry Sekulic, I'm Emma Gillespie.
In fewer than one hundred days, under sixteens in Australia will be banned from social media. Kids and teens won't be able to make a YouTube, TikTok or Facebook account. But a big question remains unanswered, how will it work. Age verification trials have been taking place over the past year, and the government has now released its findings. In today's deep Dive, we'll take you through some of the findings and look at what will happen next.
But before we get stuck into it, here's a quick word from our sponsor, Harry. This social media band story has been following us. We've been following it. It's felt like a really long time in the making, from the first time we ever heard about a potential social media ban to the legislation passing. The ban itself is not yet in place, but you've got a really interesting update for us today on the age verification stuff that's been kind of at the center of the questions around whether
or not this is going to work. But before we talk about that, let's just rewind. Can you give us a quick timeline on the big developments of late.
Yeah, this social media ban is one of those stories where in the world of journalism you say that the lead always updates, so the new headline comes through every so often. You get a new top line with every
development in this story. And I guess, going back to the very beginning's very origins, there was this really big campaign to ban social media for under sixteen's, which was mostly led by concerned parents who said that their kids were being ex to some violent and sexually explicit content, that social media was being used for bullying, that there were just some harms online that they wanted to see
more protections against. And in November last year, the government responded to some of these calls and banned social media for under sixteen's, And initially there was an exemption for YouTube. That exemption's now been overturned and that's subject to a legal challenge, which we could go into very separately in its own podcast. And the ban itself is scheduled to take place on the tenth of December later this year.
Okay, so that YouTube exemption that was reversed. We actually did chat about that on the podcast. When that happened. I think it was the end of June or the start of July off the top of my head, but will poppling to that in the show notes if you want to get up to speed on all of that debarcle and what unfolded there. But for now, Harry, we know this ban is coming into effect regardless of how
the social media platforms feel about it. But there has been a lot of confusion about the practicalities of this band. It feels like there's been a collective understanding or agreement that social media is bad for kids' mental health and we need to do more to protect kids online. But how's it going to work? Who will be responsible for enforcing this?
When I first heard that there was going to be a band for under sixteens, I was thinking that kids were going to be slugged with a fine potentially for creating an account when they're not meant to. But obviously that's not how this is going to work, and the government was pretty clear from the outset that it would be on the social media companies. They would bear the responsibility to enforce the ban. But on a practical level, there is probably a few aspects to how the band
will work. So there'll be the deterrent effect. So parents will be able to say to their kids, it's actually illegal for you to have a social media account, and that might have an effect of kids just not downloading the app or whatever it might be. They just might be dissuaded to do so. But obviously it goes beyond that, and as I said, the owners will fall on the social media companies the platforms themselves to enforce a ban, or they would risk getting a fine, which is to
the tune of about fifty million dollars. So, for example, if it was found that Meta was not taking enough reasonable steps so that's the language in the legislation, reasonable steps to prevent under sixteens from making an account on their platform, they could be slugged with a fifty million dollar fine. So the responsibility lies with them.
Okay, so the responsibilities on the platforms, not children or parents. If we've got social media companies then regulating and enforcing this, how will they and I know this is a million dollar question, how will they verify the age of their users?
Well, Australia is a first country in the world to have an age specific ban for social media. So this is kind of a novel space. This is very much new territory. And I think it's helpful to think of a nightclub. When you go to a club, you need to show proof that you are at least eighteen years old, and I think that there are some steps being taken to bring in a digital version of that for sixteen year olds on social media.
Well, Harry, it's interesting you say that, because it's one thing to be at a licensed venue in person and be able to physically present your ID, to have a bouncer or security person kind of check your ID, look you up and down, decide yep, that's you. You are free to pass. But in the digital world, in the online environment, it's just a whole different story.
And that's where we get into this space of talking about age assurance technology, which is basically verifying that you are at least sixteen years old and able to use social media. And so since the government announced its ban, it said that it was going to trial so of this technology, that it was going to get an independent review into some of the tech that's available out there. And what we have this week is the findings from
that review into existing technologies that are available. Was conducted by British firm. It was independent of the government, and it looked at forty eight different companies and the way in which they detect age assurance and how they use this technology.
Okay, so a pretty big pool I suppose of companies doing the age verification staff or building this tech. But I am really curious to know what this report found. What were the types of tech that they looked into, how did they work? Did they work?
So there were broadly three different categories of age assurance tech. So the first is called verification, and this is directly proving how old you are through official identity documents. So think of birth certificates, passports, those things that we already use day to day in our lives.
Thinking of if you are applying for a new rental and you need one hundred thousand points of IDs, that's right, You've got to find every little document that proves who you really are to get past the first hurdle on your application.
And there are some platforms that already require you to pose next to your passport photo and that way they can determine whether your fatal fitches match up to the passport itself. Okay, but also in Australia, the minimum age to get a learner's license to drive is sixteen, so that's also one way of verifying whether you are of the social media ban age if you've cleared that threshold.
Because a lot of under sixteen year olds wouldn't have on hand a lot of formal identification. Yeah, I don't know if school library cards are going to pass the test.
Well, that actually taps into some of the other areas where you can actually go through age assurance. If you've only got a school card, then some of this tech might actually think that you are of school age, maybe thirteen or fourteen years old as well, So that's one way of rooting it out. The second main age dessurance tech is called estimation, and that's where we're getting into using things like face ID to prove that you're at
least sixteen years old. So this is scanning technology that determines whether, based on your biometric analysis, whether you are at least sixteen. And the final area is called inference and this is a little bit more tricky. This is technology that uses your metadata or looks at your online digital footprint to get a sense of how old you are. This is basically taking a big overall look at how you've been conducting yourself online and whether you fall within
a likely age range. So it might just be your email domain, if you have like a work email address, that would assume that you are a little bit older. If you're on the electoral roll, you have to be at least eighteen to vote, so they will cross reference your details against that. Yeah, and there's also questions of whether they can get data from school enrollments to determine whether someone's in your ten or elevens, you know over sixteen.
So official channels because in my head, I'm thinking, if you've got someone whose browser history contains like a lot of homewares and couch options versus someone whose browser history is streaming a lot of scivity toilet, we're probably able to make a detection on their age from that.
It would be a really humbling experience with me being in my late twenties was confused for being thirteen years old on my browser You history.
That inference side is really interesting, But you mentioned estimation, and I know that this is something that had come up during the trial about you know, what is the difference between a fifteen year old's face and a sixteen year old's face. Are these kind of different categories all about working together with each other to form a more thorough picture of age verification or will platforms kind of pick and choose one.
Each Well, it's all part of what is being thrashed out at the moment, and this goes to what the legislation requires, which is the reasonable step. So they might go to these third party platforms which verify or estimate someone's age as a way of blocking anyone who's under sixteen from making an account. But also some of these
platforms already use this type of verification. Meta already uses it, so if they're harnessing it to stop people under sixteen from getting into that platforms, that might be considered a reasonable step. It depends what the E Safety Commissioner determines before the December deadline.
So, Harry, there's a big recurring question here around privacy whenever we think about the social media platforms. Privacy is obviously something that matters to users, but it's especially delicate here because we're talking about under eighteen year olds and children. Did the report say anything about that.
Yeah, it looked into some of the pretty genuine concerns about privacy breaches that could occur, especially when we're thinking about the inference technology. So that's using the online footprint to sort of build an estimation of someone's age. And the report actually warned that some continuous behavioral monitoring online
could lead to overstepping an ethical mark. So it said that there needed to be some ground rules and regulation to make sure that if the inference technology was to be used, that it is safe and that it doesn't lead to this ongoing effect.
Okay.
The other main issue that was brought up in the report was accuracy. So what if the technology accidentally starts locking people out who are over sixteen and actually able to use social media under the law.
Or similarly approves people who are under sixteen because it might think they look eighteen.
So the trials found that some of the tech could mismatch someone's age and what it called a buffer zone, and that's about two to three years. So that means that someone who's seventeen might be confused for a fourteen year old or vice versa.
Okay, so this report is acknowledging some of the shortcomings of this tech, and it's saying that there is a bit of a buffer zone. That feels kind of difficult because if it's a buffer zone of two to three years, you know, the experts are telling us that that's a critical age, those two to three years between you know, fourteen to eighteen. So I'm sure there'll be people watching that one closely.
And the language and the report says that based on its trials, the way in which the estimations have occurred are within what they call the acceptable range. Okay, so that means that there aren't enough inaccurate estimations of someone's age to cause too much alarm based on the report's findings, but it is still you know, once you roll this out on mass to every teenager and child in Australia, like that does become a bit more of a pressing
issue lasting I just note on accuracy. There was this really interesting note in the report as well about some issues with non Caucasian user is, including that First Nations peoples are underrepresented when training this data, so there could actually be an inbuilt bias to the system, which is something that we've heard about with face ID scanning technology before as well, and so the report said that there was a bit of awareness among the age assurance companies
as well that this is a shortcoming and that this is something that they need to look further into as well.
Fascinating so having more diversity across the trials to understand the data better on a more diverse range of users. We are talking, though, Harry, about a generation that's grown up online, right, We're also talking about teenagers. When you consider those two things together, you can't help but wonder is this generation that will find its way around age verification technology. Did the research consider that?
Yeah, And the standout thing to me was the use of virtual private networks which are also known as VPNs to dodge the rules. And the reason it sort of picked up my interest is because recently in the UK there was a ban on pornography websites for under eighteen's and so they rolled out this age deshurance technology and all of a sudden there was just an uptick in the number of people downloading VPNs.
And so there's a learning there for Australia exactly.
So the report basically recommended to get around the issue of VPNs, which basically makes you appear as though you're in another country where the rules don't apply, and that's where you change what's known as your IP address. But there's geolocation technology that can detect an inconsistency. So if your IP address and your regional patterns are not lined up, then this type of technology could pick up on that, okay, And it's already been used in some context as well, so it.
Is possible to identify if someone changes their VPN in Australia but has an Australian IP address. I think that that's really interesting and will probably be some relief for policymakers and parents alike to know that that exists, because if a workaround exists, we can expect teenagers to find it.
That's right. Overall, the government says that it's quite reassured by the findings that it's workable, it's scalable, and there is going to be a bit of a trial and error when rolling out the technology, but it's acknowledged there's not really a one size fits all approach that is appropriate when considering getting kids off social media. But the report does give a bit of a rundown of what
we can expect from the tenth of December. You might need to show your passport, you might need to scan your face to see if this technology picks up that you're over sixteen, or there might be an app that scans your search history, your online footprint to see how old it thinks you are. And the fact that they're still not fully sure how the technology is going to work has been picked up this week by the opposition.
So the Shadow Communications Minister, Melissa Macintosh said that the report has come ten seconds to midnight, adding that the report is not the final step by any means. The Safety Commissioner now needs to have a look at the findings and then advise the social media companies on what will form the best practice for them, and all the while the bean is just under one hundred days away. So TikTok punintended.
Sure was Thank you so much, Harry. It's so interesting because I think, you know, we hear about this ban and it might feel far away for some of us, you know, for those of us who are shockingly over sixteen. That's my secret. I'll never tell. But you know, anyone who has a log in across these platforms is hypothetically going to be affected. Right, we will, I imagine on the tenth of December, all have to verify in some way at our place on these apps.
Unless we've been on there for as long as I have, then I hope they just know that I'm well and truly.
Is not maths sing. This guy's over sixteen. Harry, thank you as always for that breakdown. We really appreciate your guidance.
Thanks Sam, I appreciate it.
And thank you for listening to today's episode. We'll be back a little later on with your evening news headlines, but until then, have a great day.
My name is Lily Madden and I'm a proud Adunda Bungelung Calcuttin woman from Gadighl Country. The Daily oz acknowledges that this podcast is recorded on the lands of the Gadighl people and pays respect to all Aboriginal and Torres Strait Island and nations. We pay our respects to the first peoples of these countries, both past and present.