Table of Contents
Rob Tyson
Rob Tyson is a director of Together, which unites people throughout the UK to uphold freedom and fundamental democratic rights, including campaigning to scrap the government’s digital ID scheme.
The government’s ‘Growing Up in the Online World’ consultation, closing end of Tuesday, is presented as a broad exercise in public engagement about children’s safety online. Yet it’s being conducted against a backdrop of unusually explicit political signalling, where ministers have already begun to articulate preferred outcomes while the consultation itself is still live. What looks like a neutral exercise in gathering opinion may instead be functioning as a mechanism for legitimising a policy direction that is already well advanced – one that extends far beyond children’s safety into broader questions of access to online services, including tighter age assurance, expanded identity verification and digital ID requirements, and emerging discussion of enforcement measures that could reach as far as privacy tools such as VPNs.
During the consultation period, the prime minister moved beyond neutral framing and into clear policy signalling. On BBC radio on April 13 he said: “We’re consulting on whether there should be a ban for under-16s.” In the same intervention he added: “The addictive scrolling mechanisms are really problematic… they need to go.” A few days later, he was reported as telling tech executives that social media “can’t go on like this” and demanding “real-world changes” while the consultation remained open. Whatever one makes of the policy intent, this is not the language of open-ended deliberation, but a direction already set.
Across the three surveys, the exercise does not simply ask what people think, it increasingly defines the terms on which thinking is permitted. The result is a questionnaire that risks producing not a neutral picture of public opinion but a carefully channelled set of responses shaped by framing, omission and the selective elevation of certain risks over others.
A consultation that starts halfway to its conclusion
At first glance, the surveys appear comprehensive. They ask about social media use, harms, benefits, parental confidence and regulatory preferences.
But the structure repeatedly narrows towards a familiar policy horizon: restriction.
Respondents are asked, in various forms:
- “Which, if any, of the following features/functions cause you to worry about your children using social media?”
- “Which, if any, of these things should be blocked for people under-16?”
- “Which of the following restrictions to children’s access to online platforms would you support?”
What is largely missing is the earlier, more fundamental question: whether this is an area where governance intervention is the appropriate starting point at all, and if so, on what basis.
Instead, respondents are placed within a pre-selected architecture in which the primary policy language is already defined: age gates, feature bans, enforcement mechanisms and access controls.
In some cases, the narrowing is not only about what is asked, but what is not imagined as a valid response category. When respondents are asked what children might be doing instead of spending time on social media, the options are overwhelmingly positive – sleep, exercise, hobbies, time with friends, family interaction. What is absent is any acknowledgement of less orderly but very real adolescent states: boredom, low mood, social withdrawal, risk-taking, or in more serious cases self-harm or criminal behaviour. The effect is not simply to measure alternatives, but to curate them.
It is not so much a consultation on options as a ranking exercise within a pre-determined policy frame.
Features defined through their worst cases
Perhaps the most striking feature of the consultation is the way neutral platform capabilities are repeatedly recast through the lens of risk.
Consider the following item: “Ability to send nude images or videos.”
On its face, this appears to be a question about messaging functionality. In practice, it is something more specific: a general communication feature is immediately defined through a highly charged edge case.
By embedding a vivid misuse scenario directly into the label of a general feature, the survey collapses two distinct concepts:
- a platform capability (sending images or video), and
- a particular form of misuse (sharing sexual imagery)
The effect is subtle but powerful. Respondents are not being asked to evaluate a system feature in the round. They are being asked to react to a vivid and emotionally charged scenario, presented as though it is the defining characteristic of that feature.
This is a classic case of what behavioural researchers would recognise as ‘salience distortion’: the amplification of emotionally striking but non-representative scenarios in ways that shape judgement about the system as a whole.
It is not that the concern is illegitimate. It is that the framing risks replacing how common something is with how vivid it feels.
Asymmetry of harm and benefit
Across all three surveys, the language of risk is not incidental but structural.
Children are asked about:
- “feeling left out”
- “worrying what others think”
- exposure to “inappropriate or harmful content”
Parents are asked about:
- children being “safer online”
- losing “control” over usage
- risks of “strangers”, “location sharing” and “AI chatbots”
And across both, harms are systematically itemised, categorised and operationalised, whereas benefits tend to appear in more generalised or open-text form.
This creates an asymmetry: risks are specified and enumerable, benefits are broad and less formally structured.
In survey terms, that difference is meaningful. Items that are listed in detail are more likely to come to mind at the point of response than those left broad or unstructured.
A pre-loaded policy direction
Perhaps the most important feature of the consultation is its underlying ‘funnel’ structure.
Respondents are repeatedly guided towards a narrow set of policy instruments:
- minimum age restrictions (often 16-plus)
- feature-level bans
- time restrictions and curfews
- enforcement and age verification systems
- platform compliance obligations
One question even asks:
- “Which platforms should be exempt from the new rules?”
Talk about jumping the gun. The existence of “new rules” is taken as given: only their scope is up for discussion. The effect is to shift participation away from whether intervention is justified and towards how it should be calibrated.
Who decides what counts as “high quality content”?
Beneath the specific design issues lies a more philosophical concern: who decides what counts as “good” or “harmful” information online?
The consultation repeatedly gestures towards harms such as:
- misinformation
- harmful content
- inappropriate material
- unsafe interaction
Yet these categories are not neutral descriptors. They depend on contested judgements about truth, value and appropriateness.
This becomes explicit in the survey’s section on “Promoting high quality content”, which defines it as material with “positive impacts on children’s learning and development” before asking: “Who would you trust to determine what is meant by ‘high quality online content’ for children 13–16?” The list of possible arbiters includes government, platform trust and safety teams, educators, youth workers, child advocacy organisations, parents, children themselves and developmental experts.
The structure of the question is revealing. It assumes that “high quality content” must be defined and authorised by designated institutions, with the only open question being who should do the defining.
This raises a deeper issue: modern regulatory frameworks increasingly operate through what might be called soft censorship mechanisms – not usually direct suppression of speech, but structural shaping of what content is surfaced, recommended or deprioritised.
In this context, questions about “safer online experiences” or “harmful content” are not simply technical. They imply that knowledge and expression require governance by institutions that define what is acceptable.
By focusing heavily on restriction-based responses, the consultation reflects the same framing: what is labelled ‘safety’ is treated primarily as something delivered through constraint, rather than through risk awareness, user autonomy or digital literacy.
The missing half of the picture
Perhaps the most revealing feature of the consultation is not what it asks, but what it does not systematically explore.
There is limited structured engagement with:
- privacy costs of enforcement technologies
- biometric age verification risks
- false positives and exclusion from services
- proportionality of interventions
- rights-based approaches to digital access
- non-restrictive safety architectures
These are not absent entirely, but they are not given the same structured prominence as restriction-based options.
The imbalance matters a great deal, because in any survey what is measured in detail is what ends up shaping the findings.
Consultation as confirmation
None of this necessarily implies bad faith. Government consultations are rarely pure exercises in open philosophical inquiry. They are instruments of policy formation and inevitably reflect institutional priorities.
But the cumulative effect here is difficult to ignore.
Across children, parents and the public as a whole, the consultation increasingly appears to:
- define the problem through risk-heavy framing
- structure answers within a restricted policy space
- embed assumptions about intervention
- amplify certain vivid but non-representative harms through design choices
The risk is not that respondents are misled in any simple sense. It is that they are gently guided towards a set of conclusions that the surveys themselves have already made easier to reach. In that sense, the consultation does not just record opinion. It helps define the boundaries of what counts as a plausible answer in a policy process that increasingly points towards the use of significant state power to reshape access to online life at scale. Participation in the consultation may still matter, if only to register dissent within a process that is already narrowing its own field of view.
This article was originally published by the Daily Sceptic.
