Okay, I have read the Age Assurance Consumer Research report (https://www.infrastructure.gov.au/department/media/publications/age-assurance-consumer-research-findings) and will do a #tljr on it as a thread.
From here I'll put the thread behind a CW and unlisted, and you can filter on #tljr.
Overall this is good research. It's nuanced and comprehensive and was done with rigour.
There are some areas where I think the report draws inferences that are not fully supported by the data.
There are some important areas where the researchers explain the data in one way with undue confidence when there are other, alternate explanations for the same data.
"These findings suggested the need for flexible, multi-faceted age assurance approaches that could accommodate diverse usage patterns while providing enhanced protection for vulnerable users, particularly in high-risk platforms and social media environments where children and adults showed divergent views on appropriate access ages."
There is no silver bullet. #tljr
BUT and this is a big but.
"The research showed minimal negative impact on intended usage, with 80.29% of adults indicating age assurance methods would either not affect or positively influence their likelihood to use websites."
I believe the best interpretation of these parts of the research is that people don't think age assurance tech will apply to them personally. They think it's only going to apply to children, or people looking at porn. #tljr
And then there's the security privacy bit.
"Analysis of age assurance methods in practice revealed significant trust and security concerns, with only 4.43% of adults fully trusting online platforms to store personal information securely, and 52.44% having experienced data breaches."
No one trusts Facebook et al to protect our data. That mistrust is based on concrete evidence and personal lived experience. #tljr
"The findings suggest that success will depend on addressing both technical and human factors, with particular attention to privacy, security, and user education."
And that, I suggest, is why it is doomed. Government has a terrible track record of implementing tech policy and is not well-placed to attempt something this complex in a hurry. If they try, they will fuck it up. There are other, less fraught options they should be exploring first, building capability and trust.
Use for non-work purposes is specifically separated from work purposes, but the "what do people do online" question weirdly doesn't include searching for information or news. 90% online shopping, 88% apps for messaging, 86% social media for personal use, 77% content streaming.
Online gaming is at 35% and, higher than I expected, online porn (both commercial and user-generated) at 28%.
People be horny. #tljr
Young people use dating apps more than old people. 18-24s at 25% compared to 2% for those 75+.
You can abandon that Tinder for Wrinklies startup idea, Dave. #tljr
"Education level also related strongly with streaming usage and peaked among Postgraduate Degree holders (85.82%)."
I wonder if that's a proxy for socioeconomics, given how much Netflix costs now. Or maybe the P in PhD stands for piracy? #tljr
"There was also considerable alignment between parent perceptions of their children’s activities online, and how the children, themselves, reported spending time online."
It turns out parents *are* engaged with their kids and *do* understand what they get up to online. Perhaps freaking out about what The Youth are doing in their bedrooms is misplaced anxiety? #tljr
However! "Less than half (45.88%) of the parents surveyed had used age-based filtering or parental controls to protect their children online. By comparison, 68.75% of children reported that their parents had set parental controls or helped them with setting rules for what they could see online."
I think the difference here is an artifact of how the question was asked of each group. Kids were asked if parents set rules for them, but parents were asked about using technical features, not if they set boundaries with their kids around screen time. #tljr
It does suggest that half of parents aren't using technical controls that already exist. It also drops with age (only 18% of 65-74 use parental controls, 53-55% of 25-34, 34-44) and women use them a 10% more than men: 51% compared to 40%).
Older people tend not to have young kids around so wouldn't need them as much.
Maybe other parents find them annoying or ineffective compared to boundary setting and non-technical measures? #tljr
There is an annoying section here that conflates "identity verification" with age assurance. 68% of adults have a myGov or NSW digital ID or something. 48% have a digital Medicare Card, 46% a digital driver license.
"When asked separately in the survey, however, if they had personally used an age assurance method online, only a quarter (24.08%) of adults confirmed that they had"
Well yeah, because myGov isn't an age assurance thing. It's an identity thing. They are different. #tljr
Anyway "Age emerged as a crucial factor in digital service adoption". 60% of those 25-34 have a digital Medicare Card (I think? It just says Medicare Card, but I think in the context they mean a digital version) and only 23% of those 75+. 57% versus 26% for digital drivers license.
Old people don't have digital ID. #tljr
Women slightly more likely to have digital ID than men (50% v 45%).
Families with kids (57-64%) more than couples (39%) or singles (39%) without kids more likely to have digital Medicare Card.
Same sort of rates (58% with kids, 40% without) for digital driver license.
Slightly more likely for employed (52%) versus not currently working (42-49%).
Important if digital ID becomes the thing you force everyone to have to access Facebook Messenger.
Where do people get their news from?
Adults: TV (57%), then social media (44%), domestic/Australian news website or app (38%), search engine (33%), radio or podcast (33%). Print news is dead at 13%.
Children 13-17 use social media (46%), then TV (42%), search engine (28%), radio or podcast (12%), domestic news website or app (10%).
I mean, kids don't watch the news as much as adults. That's not news (ahahahaha). #tljr
The report asserts that "having access to news on social media was important to all respondents" but I'd argue that when ~70% of adults say it's "Not important at all" or "Somewhat important" that means they are mostly ambivalent about having news on the socials.
Only 31% of adults saying its "very important" to have access to news on social media (23% of children 13-17) isn't something I'd advise hanging the entire news media bargaining code negotiations on. Facebook might just call your bluff.
Why do kids use social media? To be social! Surprise, surprise.
82% talking to family and friends, 75% looking at stuff related to hobbies, 67% connecting for school or work, 66% "I mostly use messaging apps", 64% creativity and inspiration, 59% connecting with people with similar interests, 54% following celebrities or influencers.
"Keeping up with news and events" is down at 50%. Would they miss it much if it wasn't there? How much are you willing to bet? #tljr
The concerns children have about using social media is very interesting.
58% Strangers contacting me
52% Spending too much time on social media
51% Cyberbullying
49% Not being able to control what shows up on my feed
47% Getting addicted to social media
44% Spending less time on physical activity
44% Spending less time with friends and family in real life
42% Comparing myself to people online
38% Seeing content that's mean for older people
4% Other
Note how mostly it's about the ability to self-manage what they do. Maybe platforms forcing everyone to watch algorithmically placed ads is a problem we should address? #tljr
No one knows who the eSafety Commissioner is.
For adults: "37.15% knew them by name only and 38.94% had not heard of them before the survey."
"By comparison, seven in ten children (71.92%) had not heard of eSafety at all"
The eSafety Commissioner might like to reflect on that a little bit, especially since they were originally the Children’s eSafety Commissioner.
We might also like to consider the difference between "activity" and "effectiveness" in this context. #tljr
"Almost eight in ten adults (78.10%) stated that they were not at all familiar with the [Online Safety Act 2021]".
Not a huge surprise, but keep that in mind whenever people go on about "there should be a law" because maybe there is one already and no one knows about it.
A reminder of how deeply weird I am for knowing anything about this stuff. #tljr
Important stat alert!
"Looking specifically at parents, reports indicated that their child had been exposed to inappropriate online content was consistently higher among:
- women (47.14%) compared to men (39.16%)
- those who resided in regional areas (58.08%) compared to capital cities (39.96%)
Bit of a mismatch in what is considered "inappropriate"? Let's interrogate this a bit. #tljr
The question actually asked "Has your child ever watched or seen any content online that was meant for or rated for someone older than they were?"
That could mean letting a 14 year old watch a movie rated M15+. That's not the same question as "have they seen something inappropriate".
The report editorialises here, and we should look out for people taking that rhetorical sleight-of-hand and running with it. #tljr
88% of parents reported that their child had not experienced cyberbullying, image-based abuse, or extortion.
11% of kids did, which is something that deserves attention, but let us keep things in perspective when we are looking at interventions that will affect *everyone*.
And importantly, don't conflate "watched a movie a bit too young" with "was extorted into sending nude pics of themselves to a pedophile that was, for once, not their uncle Jeff".
If a bad thing does happen (that 11%), parents talk to their kids (85%), and maybe the kid's school (60%). Given it might not always be something related to school, it's unsurprising that the school figure would be lower.
Police/eSafety/Other is 26-27%. There's nothing here about why that might be, so we should be careful not to infer too much. #tljr
This may alarm some: "there was broad agreement that YouTube was suitable for younger users, with 85.04% of children and 68.59% of parents indicating it was appropriate for those aged 15 years and under.
In contrast, TikTok, Snapchat, Instagram, Facebook and Facebook Messenger were considered suitable for slightly older age groups."
I'm sure Google/YouTube will be noting that whenever they're negotiating on this topic. Especially when Tiktok and Snapchat get salty. #tljr
Figure 16 on the age appropriateness of various platforms is interesting and I will include it here.
LinkedIn is seen as less appropriate for kids than Twitter/X and honestly, what sort of weird child would want to voluntarily expose themselves to LinkedIn brainworms? Apart from whoever is the new Caleb Bond, I mean. #tljr
"Understanding related strongly with willingness to engage with age verification systems"
What the report means by "understanding" appears to be whether or not they know the difference between "various age assurance methods, such as age verification, age estimation, age inference, age gating, and identity verification, by analysing their ability to match these methods to their description."
More on that in a moment.
"Online age assurance was perceived to be of great importance across all participants, commonly seen as part of a societal obligation to protect children and their development through regulation of age-appropriate content and harm minimisation."
It is interesting to see how the report frames this "harm minimisation" thing. #tljr
"Participants were asked to review an article about the introduction of age assurance technology to bar children from online pornography in Australia."
I am unable to find a citation for the specific article that was used. It does not appear to be included in the report.
"All three groups were alarmed by the statistics presented in the article on the high percentage of children exposed to online pornography, especially in pre-teen years."
It is difficult to interrogate how reliable these statistics might be, or the context in which they were presented. #tljr
@daedalus sounds like someone should fire up #ParentsForPrivacy. It won't be EFA, consider them defunct.
I'll do it through @drwaus if that was something they'd welcome, I'm doing the work anyway, with @justanotheramy mainly for regions and harm minimisation of "AI", but I'm also a bit tired.
@daedalus exactly the point, identity is not “identification”. Means of identification can be very antagonistic to identity for a not insignificant number of folk.
Age verification sounds like it’s trivial but is not & it’s also potentially a massive source of information leakage especially for the more vulnerable amongst us.
JP
Joel Michael
Sminney
ok_lyndsey