Stanford's Crusade

Policy, Passion and Restricting Access to Social Media

Introduction

The Herald’s front page for 22 July 2026 left nothing to interpretation: “I’m standing up for kids against insidious tech,” Erica Stanford declared. The timing was convenient. The article arrived hard on the heels of the paper’s own “Under the Influence” series — an unashamed, unapologetic piece of advocacy about how social media is said to be harming our youth.

Stanford has proposals to regulate young people’s use of social media. What she does not have — at least not yet — is a way through Cabinet. She cannot say when the matter will go up, nor whether she can muster the support to carry it as a Government Bill. We, for our part, do not yet know what the proposals actually are.

That uncertainty explains the outstretched hand across the aisle. And, as is customary, Chris Hipkins vacillates: Labour cannot decide whether to back the plan until it knows “exactly what it is that they’re going to propose.”

Stanford’s passion is not in doubt. Her language makes that plain:

“Tech interests are insidious and they’re everywhere.”

“They have got bottomless pits of money and I’m here standing up for our kids. I will go to anyone who’s willing to support me on this because this is what parents want.”

“They need the tools to be able to protect their kids from the harmful algorithms that they’re just defenceless against.”

This article asks three questions. What is driving the policy, and what shape might it take? Is passion a sound foundation for policy — and for the law that must follow it? And here I should confess a suspicion: the subtext of much that Stanford has said, recently and in the past, hints that the welfare of young people may not be the prime mover at all. Finally, I turn to France, where the very same instinct has just been written into law, and ask what it reveals.

The Policy Issues

Stanford’s push and Christopher Luxon’s backing both rest on a fairly specific set of stated concerns about child welfare. Yet the proposal has become entangled in far deeper questions — about digital identity and state power — that reach well beyond children.

Stanford frames the matter as a fight waged on behalf of children against the tech industry. As noted, she calls tech interests “insidious and they’re everywhere” and casts herself as “standing up for kids.”

Her concern draws on advocacy from groups such as B416, which handed Parliament a petition of some 45,000 signatures, and on the testimony of young people — one 18-year-old who delivered the petition recalled being 14 when she first encountered eating-disorder and self-harm content online. No doubt the Herald’s “Under the Influence” campaign gave that concern considerable comfort.

Notably, Stanford’s actual work has outgrown a simple age ban. Catherine Wedd’s member’s bill, which set the issue running, has been put to one side in favour of a broader programme on online harm. Her aim, then, seems less a single prohibition than a wider regulatory regime for online harm, of which age restriction is only one component.

The Prime Minister has been consistently, personally enthusiastic. He calls himself “highly supportive” and “personally invested in ensuring we deliver our own ban here in New Zealand.”

His reasoning rests on a handful of pillars. The foundation is a claimed symmetry between the physical and online worlds: “We have restrictions to keep our children safe in the physical world, but we don’t have restrictions in the virtual world — and we should.” The harms he names are consistent — bullying, harmful content, and social-media addiction, from which restricting access for under-16s would, he says, protect our kids.

The trouble is that the symmetry does not hold. In equating the two worlds Luxon overlooks — or perhaps simply fails to grasp — the nature of the changes wrought by the Digital Paradigm.

He also invokes a precedent his own government set. New Zealand has already banned smartphones in schools, and Luxon points to the “increased student engagement and improved learning outcomes” that followed as proof that regulating children’s access to technology can work.

That is a fair point. Ani O’Brien has argued for restricting access to smartphones as a way of addressing social-media harm — a proposal David Seymour then took up. In some respects it is simply the logical extension of the school-phone ban.

The other driver is the Australian model. In December 2024 Australia passed a world-first under-16 ban, and Luxon has said the Australian approach is “of interest,” fines and all, for platforms that fail to keep under-16s off their services. Were New Zealand to follow suit, the law would compel platforms to verify that users are at least 16, or face fines of up to NZ$2 million.

And there is the plain political arithmetic. A 2025 Reid Research poll found 57.8% of New Zealanders in favour of a ban and only 31.6% against. As popular, values-driven policy goes, it is close to irresistible.

The Deeper Issues — What Regulatory Model is Proposed?

Here speculation necessarily takes over. Short of the Education and Workforce Select Committee’s report, we have little idea what is actually proposed — and what emerges is genuinely contested, along lines that cut across the usual Left/Right divide. Several of the issues I have canvassed in earlier articles bear repeating.

Verifying children means verifying everyone. This is the central structural problem, and no amount of engineering removes it. As the Privacy Commissioner, Michael Webster, put it - to keep under-16s off social media, every adult must prove their age.

A law designed to protect children’s privacy would require every New Zealander to submit to identity verification simply to use the ordinary tools of communication.

Given that an InternetNZ survey found 58% of New Zealanders on Facebook daily, such a regime reaches into the everyday digital life of most of the adult population — hardly a teenage minority.

The drift toward digital ID and surveillance. A child-safety measure, critics warn, slides almost inevitably into infrastructure for identifying everyone online. Officials, reportedly unnerved by Australia’s implementation troubles, are said to be eyeing European-style models built around more centralised digital-identity and age-verification systems — state-backed machinery for online identification, rather than merely telling platforms to “ban children.”

The EU tries to soften this with zero-knowledge proofs — proving you are over 16 without revealing who you are — but even that depends on an initial, trusted step in which someone confirms you are who you claim to be.

The VPN problem. A VPN defeats an age restriction with ease, which is why VPN restrictions surfaced in official discussions — and promptly became a flashpoint.

Both Luxon and Stanford have since ruled a VPN ban out, her office insisting it was “not looking at restricting or banning VPNs.”

The reason for the sharp reaction is simple. A VPN ban would make the restrictions harder to dodge, but at a steep cost to privacy, and it would place New Zealand in company it should not want to keep — the only states that ban VPNs being China, Russia, Iran, North Korea and Belarus. The episode laid the underlying tension bare: an enforceable ban seems to demand tools that carry an authoritarian odour.

Effectiveness and unintended harm. ACT and the Greens opposed the original bill on the grounds that the rules were easily circumvented, that at-risk groups might be driven into greater isolation, and that social media harms other age groups too.

The concern for vulnerable young people is specific — the disabled, the neurodivergent, the queer, the geographically isolated — for whom online communities are lifelines, not hazards.

Academic critics add that the framing runs only one way, treating social media as something done to young people rather than a tool they use to make meaning: social scaffolding, identity formation, contact with family overseas, community support.

Beneath all of this sits the coalition problem — the one Stanford gestured at when she floated a Faustian pact with those across the aisle.

The coalition is why the policy keeps stalling. Both of National’s partners oppose it: ACT and NZ First alike say they will not support a social-media ban for under-16s. ACT’s objection is philosophical — MP Parmjeet Parmar argues that the verification required amounts to “a major shift toward online identification and surveillance” for adults as much as children, and that the party will not force New Zealanders to hand personal identification to social-media companies.

That is why National cannot pass the measure as a straightforward government bill, and why Stanford has openly courted Labour instead: “If ever there was an issue to set aside politics, it’s this one.”

The political geometry is intriguing. A National–Labour alignment would push the opposition out to the edges rather than the centre — privacy advocates, libertarians, disability advocates, queer communities and youth campaigners forming a temporary alliance, not out of shared philosophy but because they are asking the questions the mainstream debate has largely ignored.

To sum up: Stanford and Luxon are moved by a genuine and popular concern about bullying, harmful content and addiction, fortified by the Australian precedent and their own school-phone ban. But you cannot reliably keep children off these platforms without building age-verification — and potentially digital-ID — infrastructure that applies to every adult. That turns a debate about child safety into one about privacy, anonymity and state control of the internet. That is the fault line on which the whole proposal now rests.

Passion and Policy

My professional life has been given over to reasoned, objective, dispassionate argument. I approach this subject, therefore, with some hesitation — but, I hope, with balance.

Is passion a sound basis for policy? The relationship between the two is less tidy than “emotion bad, objectivity good.”

Begin with the obvious point that there is no purely objective social-media policy waiting to be discovered. The judgment that children need protecting from certain online harms, and that the state is the right body to do the protecting, is a value judgment — not a neutral fact read off the evidence. Every protective law in history — restrictions on child labour, seatbelts, tobacco advertising, alcohol age limits — was driven at least partly by moral conviction pressing against commercial resistance, and each was denounced in its day as emotive overreach.

Conviction, then, does not disqualify a minister; in a democracy it is arguably part of the job. Ministers are elected to advance values, not to administer spreadsheets. Stanford calling tech interests “insidious” is not, in itself, evidence of bad policymaking. It is a register — the public-facing, persuasive voice a politician uses to build a mandate.

The real question is narrower, and more useful. Is the conviction disciplined by evidence and open to counter-argument — or does it substitute for them and foreclose scrutiny? That is the line dividing legitimate, value-driven leadership from bad lawmaking. And there are warning signs worth watching for, in precisely the language I have flagged.

The most telling is the framing: “If ever there was an issue to set aside politics, it’s this one.”

That sentence does real rhetorical work. It casts anyone who objects — including her own coalition partners, raising substantive privacy concerns — as playing party politics rather than asking legitimate design questions.

When ACT says the select committee “failed to properly define the harm,” that is a methodological objection, not obstruction; to treat it as obstruction is exactly where moral urgency begins to crowd out the dull, essential work. Passion becomes a problem the moment it recasts reasonable dissent as bad faith. We have been down that road before.

Here is the distinction that matters most — the more so given that “law must inevitably follow.” There is a difference between the motivation for a policy and its design.

Passion is a perfectly good answer to “why do anything about this at all?” It is a hopeless answer to “what, precisely, counts as social media?”, “how do you verify age without verifying everyone?”, or “what becomes of the isolated queer or disabled teenager for whom these platforms are a lifeline?”

Those questions demand definitional precision, proportionality, and enforceability testing — the dispassionate machinery of officials, regulatory impact statements and select-committee scrutiny. Emotive conviction supplies none of it.

Draft a policy in the glow of moral clarity before the mechanics are worked out and you get precisely what has happened here, and in Australia: a popular goal colliding with an unsolved problem of implementation.

In fairness to Stanford, her conduct suggests the passion and the process are not wholly fused. She demoted the simple member’s-bill ban in favour of a wider, slower programme run through officials and Cabinet — which is arguably the machinery doing its job, slowing a fast, emotive fix.

On one reading, the passionate language is the public register while the underlying method is more cautious.

On a less charitable reading, the ban became a commitment before the hard questions were answered, and the wider “work programme” is partly a way of managing a policy announced ahead of its evidence base. Both readings are defensible on what is public.

Passion is a fine engine and a dangerous steering wheel. It is legitimate, even valuable, as the thing that forces a hard problem onto the agenda against a well-resourced industry. It becomes a poor basis for law at the point where it

(a) treats the goal’s popularity as though it settled the mechanics,

(b) recasts substantive objections as politicking, or

(c) locks in a particular mechanism before the definitional and privacy problems are solved.

The test is not whether Stanford sounds passionate. It is whether, when the Privacy Commissioner points out that protecting children requires every adult to prove their age, the response is serious engagement with that structural fact — or a retreat into the rhetoric of standing up for kids.

A Possible Subtext

Tracking the regulation of digital technology as it emerges from Australia, one trend stands out: the drive to target and control the platforms themselves. The platform is the evil to be addressed. Stanford’s recent rhetoric points the same way — insidious tech interests are the quarry. That is the subtext running beneath the otherwise laudable aim of protecting young people — an aim which, properly understood, belongs to parents rather than the State. And the platforms are the target elsewhere too, wherever access is being limited by age — in the UK, and now in France.

Let us look more closely at this issue.

Age-restriction proposals are not, in their essential logic, child-protection measures. They are instruments of state consolidation, aimed at platforms that have grown into rival centres of power. The welfare of minors is the occasion, not the cause — the one frame no one can safely oppose.

The stated rationale will not bear its own weight. If harm to adolescents were truly the driver, the legislation would be proportionate to the evidence. It is not, because the evidence cannot carry it.

The empirical picture is genuinely contested: the best-designed studies — Orben and Przybylski, for instance — find effect sizes so small they sit alongside the “harm” of wearing glasses or eating potatoes, and the causal arrow runs at least partly backwards, distressed adolescents seeking out screens rather than screens manufacturing the distress.

Governments do not, as a rule, legislate blunt prohibitions off the back of contested correlational social science. When they suddenly do, the interesting question is not the science but the appetite that went looking for it.

Notice, too, the mismatch between problem and remedy. A genuine harm-minimisation regime would be graduated, targeted and evidence-tracking — design standards, default settings, algorithmic disclosure.

What is actually being reached for is the crudest tool available: a categorical ban keyed to age. It is administratively convenient for the state and maximally intrusive on the platform, yet poorly tailored to any particular harm. The choice of instrument tells the real story.

So what is actually being built? Follow what the law requires, not what it says. An age ban is unenforceable without age verification, and age verification requires an identity-and-verification apparatus bolted onto the platform.

That apparatus is the prize. It hands the state, first, a mandated compliance surface it can inspect, license and penalise; and second, the settled principle that access to these platforms is conditional — granted and withdrawn according to criteria the state defines.

The second point is the load-bearing one. Once it is established that government may dictate who is admitted to a platform and what proof they must furnish, the hard constitutional question has already been conceded. Everything after is a matter of degree: widen the excluded category, expand the required proof, attach new conditions.

Principles, once established, generalise. The child-safety ban is the foot in the door — and doors are rarely opened just wide enough for children.

Rivals for Power

Strip away the framing and an old story remains. The platforms have accreted quasi-sovereign powers: they set the terms of public speech, mediate elections, hold the data that States covet, and can comply with a government or defy it more or less at will.

That is a rival seat of authority lodged inside the state’s own territory — and States have always moved to domesticate rival authorities, whether the Church, the guilds, the press barons or the great corporations. Bringing the platforms to heel is the structural imperative. Child protection is merely the pretext to hand.

The timing gives the game away. The regulatory push accelerated not when fresh harm data arrived, but when the platforms became politically consequential — when governments found they could not control the narrative, could not compel takedowns at will, could not count on the owners’ deference. The anxiety being legislated is the State’s, not the parent’s.

And the frame is chosen for its impregnability. Oppose a data-localisation mandate and you look like a shill – one who has an undisclosed relationship with a provider to promote its ends; oppose a “child safety” law and you look like you endanger children.

The framing launders an assertion of power through moral urgency, and pre-emptively disarms the platforms’ most effective defenders — the liberals and libertarians who would resist a naked speech-control law but cannot be seen resisting this one.

It is not that governments believe children’s welfare is at stake. It is that they have found the one lever the platforms cannot publicly grab back.

What is Happening in France?

On 21 July 2026 – a week after Bastille Day - France passed a landmark law restricting access to social media for anyone under 15 — the first European country to do so, and the latest in a wave of governments acting against what they regard as the malign effects of these services on minors. Approved by wide margins in both houses, the law could take effect as soon as 1 September. It also bans mobile phones in high schools, extending a prohibition that already covers primary and middle schools.

A personal priority of President Macron, the law is seen as a bellwether for the rest of Europe, where several countries are drafting their own. The stated motives are familiar: online bullying and coercion, exposure to extreme content, the threat of predators, and the risk to mental health.

Yet — in a moment of some irony — Macron chose social media to hail the result. “I had promised it, and now it has been voted on: social media will be banned for those under 15 starting back-to-school,” he posted the moment the National Assembly voted.

And here is the telling point. France has staked out an aggressive position on reclaiming what its officials call “technological sovereignty” from the United States and China, announcing plans to replace Zoom and other American software with home-grown alternatives.

The new law sits comfortably within that project — the more so because, in France, the most popular platforms are American and Chinese. It is hard to imagine better evidence that more is at stake here than the welfare of the young.

The mechanics follow the now-familiar pattern. Enforcement falls to the platforms, which must deploy age-verification systems to filter out underage users. There are exceptions for online encyclopedias and other services with an educational purpose.

The law provides no penalties against children or parents — and, tellingly, none that are explicit against the providers either. It is to arrive in two stages: under-15s blocked from creating new accounts from 1 September, and the ban extended to existing accounts from January 2027.

The digital minister, Anne Le Hénanff, told colleagues before the vote that the timetable was realistic “because age-verification tools already exist,” with more in development, and that the burden lay on the platforms.

Her most chilling remark, however, was this:

“For four months, all of us in France will have to prove our age. If someone is under 15, the account will be closed.”

There it is, stated without embarrassment: to keep the children out, everyone is checked.

Conclusion

The shape of the restrictions has yet to be settled. As I have argued elsewhere, the policy — restricting access by age — is one thing. The unanswered question, and the one that will largely determine public acceptance, is how it is to be implemented.

One thing is clear. As I write, ten weeks remain before Parliament rises for the election, and those weeks are punctuated by recesses in August and September. Stanford’s Crusade will almost certainly not reach the Jerusalem of legislative enactment before polling day. With luck, though, we may at least glimpse the framework.

As I have suggested, Ani O’Brien’s solution remains the neatest. Ban internet-capable smartphones for those under 16, and let them use feature phones built for the basics — voice and text — stripped of the app ecosystems, social media and web browsing of the modern handset.

It addresses the harm without conscripting every adult in the country into a scheme of identity verification. Which may, of course, be precisely why it will be passed over.