When Adolescence landed on screens earlier this year, it did something rare for prestige television: it made parents sit in silence after the credits rolled. The series — a taut, grief-laden drama about a teenage boy drawn into violence through online radicalisation — arrived with critical praise and a cultural weight that felt overdue. But for all its insight, Adolescence was a dramatisation. The reality unfolding on platforms like Telegram, right now, is considerably harder to watch.
We are living through a period of extraordinary digital exposure for children. The average child in a high-income country receives their first smartphone before the age of twelve. By fourteen, they are navigating a media environment so fragmented, so algorithmically sophisticated, and so deliberately opaque that most adults cannot follow them into it — let alone protect them once they are there.
The Telegram Problem Is Not What You Think
Telegram is not, in itself, a dangerous application. For journalists, activists, and citizens in authoritarian states, its encrypted architecture is a lifeline. But that same architecture — minimal moderation, anonymous channel creation, no content audit trail — makes it extraordinarily hospitable to bad actors seeking influence over vulnerable minds.
The issue is not simply that harmful content exists on Telegram. Harmful content exists everywhere. The issue is that on Telegram, harmful content is delivered without friction, without counter-narrative, and — crucially — without any record that it ever happened. A child who stumbles into a channel promoting self-harm, extreme ideology, or criminal behaviour is not encountering a single bad post. They are entering an environment engineered to feel like consensus.
When a platform records nothing and moderates nothing, it does not simply host harm — it industrialises it. The breadcrumbs of radicalisation vanish before anyone thinks to look.— GlamBon Culture Editorial
Bots, Amplification, and the Architecture of Influence
To understand what is happening to children in unmoderated spaces, it helps to look at a historical case study in how digital manipulation at scale actually works. The 2016 US presidential election — documented in detail in films like The Great Hack — offered the world its first clear view of psychographic targeting as a political weapon.
The method was not sophisticated in the way a cyberattack is sophisticated. It was sophisticated the way a con is sophisticated: it understood people. Millions of data points were used to identify individuals whose beliefs and behaviours sat in a zone of instability — those who could be nudged. Once identified, these individuals were served a curated, relentless stream of content designed not to inform but to push. The platform, in that case, was Facebook. The technique was micro-targeting. The result was a measurable shift in political behaviour among a demographic that believed they were simply reading the news.
That blueprint has not been retired. It has been refined, decentralised, and pointed at a far more vulnerable audience.
On mainstream platforms, even poorly enforced community standards create friction. A post can be reported. An account can be flagged. A harmful trend becomes visible to moderators — however slowly they respond. On Telegram, a closed group can operate for years without any external visibility. There is no equivalent of a content moderation queue. There is no data trail for researchers or regulators to subpoena. The harm happens in a sealed room.
How the echo forms
A child who joins a Telegram group for what seems like a shared interest — gaming, fitness, anime, music — may encounter no harmful content for weeks. This is not accidental. Grooming, whether by a human actor or a coordinated bot network, tends to begin in the affirmative. The community validates. It feels warm. It feels like belonging at an age when belonging is the central psychological project.
Then the framing shifts. Slowly, incrementally, the content takes on a harder edge. Outgroup hostility. Glorification of transgression. Content that frames severe behaviour — violence, self-harm, criminal acts — as either romantic, inevitable, or justified. Because the platform tracks nothing and removes nothing, there is no external check on this progression. The child's only reference point is the community itself — which has been shaped, often deliberately, to make the extreme feel normal.
The Case for a Verification-First Internet
There is a version of this conversation that devolves quickly into surveillance anxiety — the fear that digital identity verification means governments or corporations watching everything. That concern is not entirely misplaced, and it deserves serious engagement. But it should not be allowed to crowd out a simpler, more urgent question: why, in 2025, can a ten-year-old join a Telegram channel with no more friction than pressing a button?
The current model for age gatekeeping online is largely an honour system. Platforms ask users to confirm they meet an age requirement. Users — including millions of children — confirm they do. The platform records the confirmation. No one checks. The door is open.
Robust digital identity management does not require surveillance. It requires a shift in the technical architecture of onboarding: systems that can verify a user's age through cryptographic or biometric means, without retaining the underlying personal data. The verification event is confirmed; the data behind it is not stored. A child cannot gain entry. A parent does not need to be present. No one is watching. But the door is locked.
The honour system is not a policy. It is an abdication. We do not let children walk into a bar because they say they are adults. The logic that exempts platforms from the same standard has always been commercial, not ethical.— GlamBon Culture Editorial
Reclaiming the Narrative
None of this is an argument against the internet. It is not an argument against privacy, encryption, or the free flow of information. It is an argument that the infrastructure of digital life was built without adequate consideration of its most vulnerable users — and that the cost of that oversight is now arriving in the form of damaged young people, grieving families, and crimes that trace their origins to a sealed chat room.
Adolescence was a warning dressed as entertainment. The series did what journalism often struggles to do: it made the abstract personal, the statistical human. But warnings are only useful if they produce action. The action required here is not more content moderation — though that too is needed. It is a fundamental rethinking of who gets to enter which spaces, and how that boundary is enforced.
We have the technical tools. We have the policy frameworks, in draft form, in legislatures across Europe, Australia, and the United States. What has been lacking is the collective will to prioritise the safety of children over the growth metrics of platforms. That is a cultural problem as much as a technical one. And culture, at least in part, is where it gets solved.