CHAT CONTROL III. – The Road to a Fully Identified Internet, Worldwide

The Advertiser Precedent

Since the mid-2010s, major advertising platforms have been able to construct remarkably precise profiles of individual users.

By combining search history, likes, shares, location data, device fingerprints, browsing patterns across sites, and behavioural signals, companies could infer hobbies, political leanings, religious interests, and even sexual orientation with high accuracy.

These profiles powered micro-targeted advertising campaigns.

A 2015 research paper demonstrated that Google’s targeted-advertising ecosystem could be exploited to infer personal information about users by combining Google advertising tools with information collected through websites in the Google Display Network.

What Chat Control-style rules ADD is the legal obligation and central coordination to apply similar analysis at scale, in for purposes defined by the state.

A Worldwide Pattern, Not an EU-Only Story

Age verification requirements, content risk assessments, and restrictions on younger users are advancing globally.

Australia implemented a social media ban for under-16s.

Various US states have passed age-verification laws for certain platforms and content.

The European Union is developing age-verification tools linked to digital identity systems.

The United Kingdom has moved furthest in turning these obligations into everyday restrictions on popular services.

This is not a solely European experiment.

Gaming and Everyday Restrictions: The UK Example

In the United Kingdom, rules connected to the Online Safety Act now include tighter controls on under-16s in gaming environments.

From spring 2027, platforms are expected to restrict stranger contact and livestreaming features for younger users on services such as Minecraft, Fortnite, Roblox, and Xbox.

Age checks and default safety settings are required for "high-risk" social functions. Yes, like chat.

These restrictions are difficult to enforce effectively in practice.

Children (and potential offenders) can use services to appear in other jurisdictions, lie about their age during account creation, fool the AI with another AI or migrate to less regulated platforms and servers.

Enforcement therefore often relies on imperfect technical measures that determined users can circumvent.

Digital ID and the end of practical anonymity

Age verification does not logically require every website to know a person’s full legal identity. Privacy-preserving systems can theoretically prove only that someone is over a threshold, such as 16 or 18, without disclosing their name or complete identity.

The EU describes its current age-verification application - hacked in under 2 minutes - as a means of proving age without directly sharing unnecessary personal details with platforms. ;D

Poor implementation could make sessions linkable, exclude users, produce records of verification events or gradually expand to “prove exactly who you are.”

The worst case is an internet where anonymous reading, discussion and publication become impossible. Whistleblowers, critics, political dissidents, abuse victims, people researching sensitive health matters and anyone living under a hostile family or government would bear the cost first.

People also use the internet to discuss politics, organise demonstrations, circulate petitions, coordinate campaigns and build lawful opposition movements. If every account, message or verification token can be linked to a real identity, authorities gain the ability to map those networks without anyone committing a crime.

Once the identities of its organisers, donors and participants are known, online activity can produce consequences in the physical world: police visits, employment problems, account closures, financial scrutiny, travel restrictions or pressure on family members.

The result would not only be more direct repression. It would also produce widespread self-censorship.

People would avoid controversial discussions, distance themselves from lawful campaigns and remain silent on political or social issues because they could no longer be certain who was recording, analysing or storing their participation.

This is not a theoretical concern invented for the distant future. These systems are already being applied in all social media algorithms.

Removing anonymity would make that process cheaper, faster and more comprehensive.

The noob & stale state

This debate also assumes that the state is competent enough to supervise children’s online lives better than their parents. That assumption deserves far more scrutiny than it receives.

Children do not become safe online because a platform hides a feature from them or because an authority forces everyone through an age-verification system. They become safer by learning how the internet works, recognising manipulation, understanding scams, dealing with abuse and knowing when to ask for help.

That requires trust, communication and experience.

A child who is protected from every unpleasant interaction until adulthood does not necessarily become safer.

It may simply enter the internet later, with less experience and less confidence, while the scammers, abusers and manipulators remain exactly where they were.

Parents who speak openly with their children, set clear expectations and remain involved can teach judgement in a way no regulator can. A government cannot know the maturity of an individual child, the trust within a family or the rules that parents have established.

It can only impose broad restrictions on millions of people at once.

Those restrictions will inevitably be designed by officials, regulators and contractors who often understand neither the technology nor the families they are attempting to govern.

The state may punish crimes and pursue actual offenders. It should not replace parents, dictate how every child must use the internet or force adults to surrender their privacy because some families may need help managing online risks.

The goal should not be to produce children who have never encountered anything difficult online. It should be to raise young adults who know how to recognise danger, respond to it and ask for help when necessary.

Potential damage

There is privacy damage from handing identity documents, facial data or behavioural profiles to verification providers.

There is security damage from creating valuable databases connecting identities, ages, accounts, browsing activity and profiling, accessible to the state authorities.

There is exclusion damage when people without suitable documents, cameras, smartphones or successful facial-estimation results cannot access services.

There is speech damage because anonymity protects lawful but unpopular expression.

There is competition damage because major platforms can absorb regulatory costs more easily than small forums, independent developers and open-source services.

There is childhood-development damage if children are cut off from communities, creative tools, support networks and opportunities to learn how to navigate risk under guidance.

There is displacement damage because children and offenders may migrate toward less regulated, less visible and more dangerous services.

Australia’s experience already demonstrates the basic enforcement problem. Researchers found widespread failures in initial age checks, while studies involving young users found that many regarded restrictions as unfair or ineffective and learned how to circumvent them.

The Fundamental Questions

These developments raise questions that go beyond technical implementation.

Is heavily restricting or gating children’s access to large parts of the internet the most effective way to protect them, or does it simply push this activity into less visible spaces?

Is mass analysis of all communications a necessary or proportionate tool for preventing grooming and exposure to harmful material?

And most basically: should authorities concentrate resources on investigating and responding to crimes that are occurring or have occurred, rather than building systems designed to monitor potential future behaviour across the entire population?