This post is something I’m working on to support my Masters of Teaching students to grapple with the diverse public discussion happening right now in Australia that impacts educators. We start with a specific pin in time, the Social Media Age Ban implementation and subsequent ripple effects. By now that debate has its camps, its talking points, and its exhaustion, and I don’t think one more position piece weighing in on “right or wrong” adds anything. This is a companion piece to a professional development session I’m running for pre-service and in-service teachers in the School of Education at La Trobe, and it exists to do something else: to walk through what’s actually happened in Australian digital policy over the past few weeks — a new duty of care bill, a live PISA scandal, a resurgent “brain rot” discourse, a quietly abandoned decade of screen-time guidance — and use all of it to notice a pattern sitting underneath the noise. The pattern is this: children’s attention has become the economic product that platforms, governments, schools, parents, and even researchers are all bidding for at once, mostly without coordinating with each other, and mostly without asking the one question that should come first. Not how do we manage this resource, but who is this person becoming, and what do they need from us to get there.

Kids are not robots. That sentence sounds almost too obvious to write down, and yet very little of what I’m about to walk through was actually built with it fully in mind.

Naming the market

Let’s start with the plainest possible version of the claim: attention is a finite, biological resource, and right now there are at least five distinct institutions competing for a young person’s share of it, most of them unaware they’re even in the same market.

The platform is the most obvious bidder, and the most honest about it, because its business model depends on it. This isn’t speculation — it’s been demonstrated experimentally. In the 2020 Facebook and Instagram Election Study, a rare collaboration between Meta and seventeen independent academics who retained final say over publication, researchers switched a subset of users onto reverse-chronological feeds for three months. Those users spent measurably less time on the platforms afterward — the researchers’ own conclusion was that the algorithmic feed had made the platforms “more compelling.” That’s not a moral judgment about Facebook. It’s a controlled experiment showing that engagement is engineered, not incidental.

The school is a second bidder, and it comes to the table wanting something quite different: a legible, rankable outcome. PISA scores, NAPLAN results, a number that can be reported to a minister and compared against last year’s number.

The government is a third, and what it wants is something announceable — a bill, a ban, a toggle, a press conference. I’ve written elsewhere about how thin some of that turns out to be once you read the actual clause rather than the press release, but the incentive structure is consistent: legislatures can mandate an age check. They cannot mandate a child’s capacity for judgment into existence with a clause and a penalty regime.

The parent is a fourth bidder, and the most recent EU Kids Online report — 28,465 children surveyed across 19 European countries — shows exactly which way parents have been reaching. Technical and restrictive tools have risen sharply since 2020: content blocking up from 22% to 38% of families, activity tracking up from 21% to 40%, GPS location tracking up from 15% to 42%. Over the same period, the proportion of children who say their parents talk to them about what they do online fell from 66% to 50%. Parents haven’t stopped caring. They’ve shifted from a conversation to a control, because a control is the only lever that feels like it’s actually doing something.

And then there’s research itself, which I include not to be cute but because it’s a genuine fifth claimant. Kaye, Orben, Ellis, Hunter and Houghton’s methodological critique argues that “screen time” (a ‘protaginist’ in this story) as a construct is conceptually incoherent — no agreed definition, wildly different activities bundled under one word, self-report accuracy as low as 34% in some studies. Yee and Luyt go further, in a 95-family Singaporean ethnography, asking bluntly who benefits from the screen-time frame surviving despite its incoherence: platforms, because it deflects blame from design onto usage; governments, because it’s cheap, visible action that doesn’t require regulating an industry; and researchers, because new constructs are what get funded. They call the industry’s own screen-time tracking tools “regulatory theatre” — the same phrase Amanda Third used, entirely independently, to describe Australia’s social media summit process.

None of these five bidders is overtly malicious. Most of them believe, correctly, that they’re acting in a young person’s interest. But none of them is coordinating with the other four, and not one of them, in the entire policy landscape I’ve spent the past month reading, opens with Biesta’s question rather than their own.

What the bidding has displaced

Gert Biesta’s theorisation of education’s purpose — explored recently by Juvonen, Huilla, Kosunen, Thrupp and Toom through ethnographic work in two Finnish primary schools — names three domains a school (and, I’d argue, any adult responsible for a young person) is meant to hold in balance. Qualification is the transmission of knowledge and skills — measurable, testable, comparable. Socialization is learning to be within a culture — its norms, its unwritten rules, absorbed as much as taught. Subjectification is becoming a person capable of independent judgment and action — not just knowing the rules or fitting in, but having the standing to question or resist when it matters. Biesta’s own phrase for it is “qualified freedom”: freedom paired with responsibility, not freedom alone.

The paper’s central warning, borne out in the ethnographic data, is that qualification’s logic keeps colonising the other two domains, because it’s the only one that’s measurable, and measurability is what a system under pressure to produce legible outcomes will always reach for first. A pupil’s growing capacity for independent judgment gets evaluated as if it were a spelling test — right or wrong, on schedule or behind — and something real is lost in the translation.

I don’t think this is a Finnish classroom problem. I think it’s the exact mechanism running underneath every piece of Australian, EU, and UK digital policy I’ve read this year. The American Academy of Pediatrics retired its decade-old fixed-hour screen-time limits in January 2026, replacing them with the “5 Cs” — Child, Content, Calm, Crowding out, Communication — because, in co-author Libby Milkovich’s own words, “people really want the concrete, easy advice, and trying to communicate the nuance is really difficult.”

That’s an official body admitting, on the record, that a number was chosen for its legibility, not its accuracy. Meanwhile Australia’s under-16 ban, the digital duty of care’s opt-out toggle, Victoria’s flat two-hour school screen-time cap, and the UK’s spring 2027 package are all, structurally, the same kind of answer: a verifiable threshold, an audit trail, a yes/no. What’s consistently absent (present, at most, as a single unfunded line item) is anything aimed at socialisation or subjectification.

The ARC Centre for the Digital Child’s Manifesto for a Better Children’s Internet names this directly, among its seventeen principles: “a move away from the over-reliance on ‘parental controls’ as the solution,” and “less focus on protecting children from the digital environment and more focus on protecting them within it.” That second line is, I think, the cleanest possible statement of the category error running through the entire policy landscape. Exclusion-then-access is a qualification-shaped fix. The actual problem was always socialisation and subjectification-shaped, and no amount of tuning the age dial turns one into the other.

Watching it happen this week

If this sounds abstract, it wasn’t abstract this week. Australia’s 2025 PISA results landed on 9 September, and four different, well-credentialled people read four different causal stories out of the same dataset within about seventy-two hours of each other.

Education Minister Jason Clare read screen time: “These results and Pisa over the past show us students who use screens a lot score lower than those who don’t,” he told Sunrise, calling for “a digital detox” in Australian schools. Blaise Joseph of the Centre for Independent Studies read classroom disruption: “Students cannot learn effectively when classrooms are regularly disrupted,” he told multiple outlets — worth knowing that CIS had already published a dedicated paper on classroom behaviour before this news cycle broke, so his PISA commentary is amplifying an existing institutional position rather than forming a fresh judgment from the data. Jonas Bertling, ACER’s research director, who actually administers the test, gave the most hedged account of the three: “the disciplinary climate also still needs attention… we’re really standing high, but I think we have some strengths to build on” — multi-causal, resourcing-focused, declining to pick a single driver.

And then there was Therese Hopfenbeck, Professor of Educational Assessment at the University of Melbourne and, not incidentally, chair of the OECD’s own PISA 2025 and 2029 questionnaire expert committee — meaning she has more authority to interpret what this specific instrument can and can’t tell us than anyone else quoted that week, and no single policy outcome riding on the answer. Her verdict, in The Conversation, two days after Clare’s press conference: “The PISA survey doesn’t report on the reasons for these trends… The OECD has also suggested digital devices could have made an impact on global reading skills, but this is only a suggestion at this point, and more research is needed.” She went further, flagging that even the headline distraction figures — 41% of Australian students reporting noise and disorder, against a 31% global average; 39% reporting peer distraction by digital devices, against a 28% OECD average — are self-reported perception data, not observed classroom time: “What some students report as disturbances might be productive discussions for other students.”

Here’s the sharpest fact in the whole case study, and it comes from the Guardian’s own reporting, not from me: Australia’s public schools have had mobile phone bans in place, state by state, since as early as 2020 — Victoria, WA, Tasmania and NSW primary schools that year, NSW high schools from 2023, Queensland’s full statewide ban, covering breaktimes and smartwatches, from 2024. By the time the 2025 PISA test was sat, most Australian students had spent the majority of their secondary schooling under some form of phone restriction. Distraction is still eleven points above the OECD average. The qualification-shaped fix — the thing you can legislate, measure, and announce — has been in place for years, not months, and the number still hasn’t moved the way the theory predicted.

I don’t think Clare or Joseph are acting in bad faith. I think they’re doing exactly what qualification-logic policymaking always does: reaching for the single, legible, actionable cause, because that’s the only kind of cause a press conference or a think-tank paper can actually work with. Hopfenbeck’s job, and her only professional stake in the outcome, is getting the interpretation of the instrument right. That’s precisely why she’s the only one of the four saying, plainly, that we don’t yet know.

Reframing what kids are actually doing

This is where I want to slow down, because I think the discourse around “brain rot” has been doing real harm to how adults — including, sometimes, teachers — read what’s actually happening in front of them.

Anna Götzfried and Maxi Heitmayer’s study, twenty-four semi-structured interviews with UK and German participants aged 13 to 26, found something that inverts the standard causal story. The “brain-rot feeling” — fogged out, low-selectivity, autopilot — precedes the content, not the reverse. Their participants described a mental state produced by sustained platform engagement, for which brainrot content was a low-cost response, not the cause of the state itself. They coined a term for this: anti-gratification — the deliberate seeking-out of content that makes no demand and offers nothing, precisely because everything else in a saturated media environment is already demanding something. One participant, P19, put it as sharply as any policy paper I’ve read this year: “If everything is content now, then brainrot is anti-content. It’s the only way to be free online, which is to do something so stupid it can’t be monetized.”

And then there’s P21, whose full quote I want to give in full, because I think it’s the single most important line in this entire research trail:

“Before the government regulated the tobacco industry, everyone was smoking, even though they knew it was bad. Because the environment encouraged and allowed for that to happen… Now we need to get into a state where the environment shifts so young people’s habits shift.”

Read that again. That is not a kid asking an adult to rescue them from their own weak willpower. That is a kid correctly, precisely diagnosing that the fix isn’t located in their own behaviour — it’s located in the design of an environment that five uncoordinated institutions are simultaneously extracting from. When we pathologise brain rot as evidence of decline, we’re mistaking a coping mechanism for a malfunction. It’s closer to triage: a young person refusing to hand over any more of an already-contested resource than they absolutely have to.

This is also, I think, the right way to understand what teachers describe as a trisector of challenge — distraction, behavioural disruption, and disengagement — not as three separate classroom-management problems each requiring its own policy, but as three visible symptoms radiating from one shared, shrinking resource: a recalibrated attentional baseline. Distraction is attention pulled away in the moment. Disruption is that frustration expressed outward. Disengagement — boredom — is the baseline itself having reset, so that ordinary, unstimulating pedagogical pace now reads as unbearable by comparison to an environment engineered to never let a gap in stimulation open up. A phone ban can plausibly reduce the first. It does nothing to restore the second or third, because neither was ever a function of device presence alone — they’re a function of a capacity that never had the chance to develop, because nothing in the environment around a young person, including some of what schools themselves now do in the name of “engagement,” has been interested in protecting the conditions under which it could.

What actually builds the capacity

So if bans and controls don’t build it, and if brain rot is a coping response rather than a deficit, what does the evidence actually point toward?

The EU Kids Online data is unusually direct on this. Enabling mediation — conversation, co-use, shared exploration — correlates with more digital skills and more opportunities, alongside more risk exposure. Restrictive and technical mediation correlates with fewer risks, but also fewer skills and fewer opportunities. The report’s own researchers are honest that the combination everyone actually wants — more capability, less harm — “remains elusive.” Nobody has found it yet. But the data also shows something more specific and more useful than that admission alone: the single most common positive parental behaviour recorded wasn’t a rule or a filter, it was “listening when children wish to renegotiate rules.” That’s the empirical signature of what I’d call accountable freedom in practice — not a fixed control that lifts entirely at a set age, but an ongoing, responsive negotiation where demonstrated trust expands what a young person is given latitude to do, continuously, the way a driving instructor extends independence as competence is shown rather than all at once when a licence arrives.

The ARC Centre for the Digital Child’s Manifesto for a Children’s Internet, converges on the same structural answer from a policy-design angle rather than a parenting one. Its clearest, most quotable principle — “less focus on protecting children from the digital environment and more focus on protecting them within it” — names precisely the category error running through the ban, the duty of care’s opt-out toggle, and Victoria’s screen-time cap alike. Its call for regulation that “strikes a fair balance between government policy, technology company policies, and personal responsibility” is a three-way structure, not a binary choice between control and freedom — and it points to something with real regulatory teeth that doesn’t run through surveillance at all: design codes.

The UK’s Age Appropriate Design Code, Ireland’s Fundamentals for a Child-Oriented Approach to Data Processing, France’s Les droits numériques des mineurs — these place proactive, audited obligation on platforms to build safety into products before deployment, rather than asking parents or regulators to catch harm after the fact. Australia’s own eSafety Commissioner has a Safety by Design framework in the same family, but it remains voluntary — a genuinely weaker instrument than its European counterparts, and a real, specific gap in what Australia has actually built compared to what the research consensus says the better model looks like.

None of this, on its own, is something a single teacher controls. But some of it is, and I think this is where the session I’ve built around this material needs to land. A teacher is not meant to be a single point of failure or a single point of control — young people raise themselves through the whole ambient network of relationships available to them: peers, other trusted adults, family, school, and yes, increasingly, the platforms themselves, which are now a permanent, uninvited member of that network with commercial incentives. What a teacher actually has leverage over is smaller than a policy, but it’s real: availability without the automatic cost of surveillance, since a young person discloses less to an adult who is monitoring more, not more; scaffolding that shrinks as competence is demonstrated rather than dropping away entirely at one fixed age; treating a recoverable mistake as a teaching moment inside a relationship rather than a logged event that triggers tighter control next time; and, above everything else, protecting slow, undistracted time as a pedagogical value in its own right — not because it makes a lesson more competitive with a phone, but because sustained, unstimulated attention held over time is the one precondition subjectification actually requires, and it is the single thing that none of the five bidders competing for a young person’s attention has any institutional incentive to protect.

Where this leaves us

I opened by saying this isn’t an argument about whether the ban is right or wrong, and I want to close the same way. Reasonable, well-informed people in the room I’ll be running this session for will land in different places on that question, and that’s fine — it’s not what this piece, or that session, is trying to resolve.

What I do want to leave you with is this: nearly everything currently funded, legislated, and announced around young people’s digital lives — the ban, the toggle, the flat-hour cap, even some of what schools do in the name of student engagement — is a qualification-shaped answer, chosen because it’s the only kind of answer a bill or a press release can actually contain. The evidence, gathered from an EU study of 28,465 children, an Australian research centre’s manifesto, a peak US pediatric body’s own retreat from the metric it invented, and — this week — the actual PISA data everyone’s currently arguing about, points somewhere else entirely: toward relationship, toward scaffolding that responds to demonstrated capability rather than age, toward protected time, toward design accountability that doesn’t run through surveillance. None of that fits neatly into a headline. All of it is slower, harder to measure, and much closer to what education was actually for in the first place.

Kids are not robots, and the thing everyone currently claims to be protecting them from — or extracting from them, more honestly — is a resource that only ever becomes a person’s own through exactly the kind of attention none of us can legislate into existence. We can only build it, one room, one relationship, one protected hour at a time.