Protecting Our Kids or Profiling Them? Why Privacy Concerns Specific To Canadian Bill C-34 Are So Important For Parents To Understand


CAVEAT: Here’s the challenge with age gating, you need a system to prove your age which in itself creates real risk for all of us. This article distinguishes documented technologies and regulatory provisions from potential future privacy risks. It does not establish that Canadian age gating legislation mandates behavioural biometrics or that every age assurance company engages in cross platform surveillance. The article reflects publicly available information as of November 15, 2026. As Canada considers restricting social media accounts for children under 16, an important question deserves more public attention and explanation that can be understood by parents and caregivers, “What information will companies need to collect about our children, and potentially every Canadian, to enforce these restrictions under Bill C-34?” Again, we agree with much of Bill C-34, but specific to the age gate of 16 we are not in agreement, and this article explains the reason “why” we should all be concerned.
At The White Hatter, we have long advocated for a safer online world for children and teens. We believe technology companies should be held accountable for how their products are designed, how they collect and use personal information, and how they respond when young people experience harm. However, we also believe that legislation introduced in the name of protecting children deserve the same scrutiny we apply to the technology companies those legislative policies seek to regulate.
Canada's proposed Bill C-34, the Safe Social Media Act, introduces a particularly important question (1). The legislation would establish a minimum age of 16 for accounts on certain social media services, supported by age verification or age estimation measures. It would also introduce broader safety obligations and permit exemptions for services that demonstrate sufficient protections for children.
On the surface, requiring someone to prove they are old enough to access a social media platform may seem straightforward. After all, we already require proof of age for certain products and activities in the physical world such as driving a vehicle or purchasing alcohol. However, verifying someone's age on the internet is fundamentally different from asking a person to showing your drivers licence to a police officers or purchasing alcohol at a store.
When someone shows a cashier their identification, the cashier generally needs to establish one fact, whether the customer meets the legal age requirement. The cashier does not need to record how the customer walks, how quickly they speak, which stores they previously visited, what they purchased last week, or how they interact with their phone. In the digital world, however, some technologies can collect and analyze precisely these kinds of behavioural signals.
This brings us to an emerging area of cybersecurity and digital identity known as behavioural biometrics, along with related technologies described as continuous authentication, age inference, and risk based identity verification which are not often discussed or even understood with parents and caregiver. These technologies have legitimate security applications, including detecting fraud and preventing account takeovers. However, their potential use in age assurance systems raises important questions about privacy, consent, commercial incentives, and the amount of information companies should be permitted to collect about Canadians, especially when it comes to our kids.
The issue is not that every age verification system uses behavioural surveillance, the fact is many do not. Rather, the concern is whether increasingly demanding age verification requirements could encourage a digital identity industry in which collecting more personal information becomes commercially attractive. We believe parents, caregivers, and educators should understand this distinction before accepting assurances that age verification is simply a technical solution to keeping children safe.
What exactly are behavioural biometrics?
Most of us are familiar with traditional biometrics. When you unlock your smartphone using your fingerprint or facial recognition, the device is using a physical characteristic to help establish your identity, however, behavioural biometrics works differently. Instead of examining what your body looks like, it examines patterns in how you behave.
For example, a behavioural biometric system might analyze how quickly you type, how long you pause between keystrokes, how you move a computer mouse, the pressure or rhythm of your touchscreen interactions, or how you physically handle your smartphone. Over time, certain combinations of these behaviours can help a system distinguish one user from another.
Imagine that someone steals your online banking password and successfully signs into your account. The password is correct, but the person using it moves through the website differently, types with an unfamiliar rhythm, and interacts with the device in ways that differ from your usual behaviour. A behavioural biometric system might recognize those differences and flag the session for additional security checks. This is a legitimate and potentially valuable application of the technology.
It is also important to understand that behavioural biometrics does not necessarily mean a company knows everything about you. Some systems examine a limited collection of interaction signals without recording the actual words typed or retaining a detailed personal history. The privacy implications depend on what information is collected, how it is processed, whether it can identify someone, how long it is retained, and whether it is combined with other information. That distinction becomes especially important when we begin discussing children.
Continuous authentication: When proving who you are never really stops
Traditionally, authentication happened at a particular moment. You entered a password, used a security key, or unlocked your device with a fingerprint. Once the system accepted your credentials, you were allowed to continue. Continuous authentication changes that model. Instead of checking your identity only when you sign in, a system can continue evaluating signals throughout your session to determine whether your activity remains consistent with the legitimate account holder. Think of the difference between showing a ticket when entering a concert and having security continually assess whether you belong there while you move around the venue.
In cybersecurity, continuous authentication can help detect situations in which someone gains unauthorized access after a legitimate user has already signed in. However, when ongoing verification involves behavioural profiling, it also creates the possibility that monitoring becomes a routine condition of accessing digital services.
This is where the distinction between security and surveillance becomes important. A system that analyses a few interaction signals locally on your device, discards them after processing, and only alerts a service when something suspicious occurs is very different from a system that continuously collects behavioural information, links it across websites, and maintains a persistent profile of your activities. Both might be described using similar marketing language, but their privacy implications are substantially different.
The question for parents and caregivers is not simply whether continuous authentication works, it’s whether the amount of information collected is proportionate to the problem it is intended to solve.
The important connection to age verification: How old does the internet think you are?
There is another technology that parents and caregivers should understand, known as age inference. Age inference is different from traditional identity verification. Instead of asking someone to prove their date of birth using identification, a system attempts to estimate their age based on information already available about them.
Imagine a 14 year old who signs up for a social media account and enters a false birthdate claiming to be 19. A platform might examine other signals that suggest the user is younger. These could include the types of accounts they follow, their interactions with other users, the content they engage with, or patterns associated with school-aged users. An AI model could then estimate that the account is more likely to belong to someone under 16, and this is not merely a hypothetical possibility. Australia's privacy regulator explicitly identifies online behavioural patterns, social connections, group memberships, and activity associated with school schedules as possible age estimation signals.
This is where the relationship between age verification and behavioural profiling becomes clearer. A platform may not need to know exactly who a child is to make an educated guess about their age. It may instead use statistical patterns to estimate whether the account belongs to someone in a particular age group.
However, the system's ability to make that prediction depends on the information available to it. The more signals a system can examine, the more opportunities it may have to improve its predictions, although additional data does not automatically guarantee greater accuracy. That creates an important policy question, “Should a company be permitted to monitor and analyze a child's online behaviour to determine whether that child should be permitted to use the service in the first place?”
There is also a technical limitation worth understanding. Behaviour based age inference generally requires a history of activity. It cannot reliably determine the age of a completely new user if it has no relevant information about that person.
In October 2026, the UK's communications regulator, Ofcom, specifically concluded that existing behavioural age-inference methods were not suitable as a standalone gate for preventing under-16s from creating accounts. However, Ofcom recognized that these systems could complement other age checks by identifying people who may have circumvented restrictions (2). This matters because it suggests a possible two stage approach, verify a person's age when they first register, then use additional signals to identify accounts that may have bypassed the restriction.
Whether Canada ultimately permits, requires, or limits such an approach will depend on the regulations and technical standards adopted under its legislation which have we have yet to see, but our Canadian government are trying to pass in Bill C-34.
How age verification systems actually work
Not every age verification method requires the same information. Some involve sensitive documents or facial analysis, while others can provide proof of an age threshold without revealing a person's identity.
The first method involves verifying a date of birth against a trusted document, such as a passport. The second uses facial age estimation, where software examines a person's face and estimates their age or age range. The third uses a digital credential, which may allow someone to prove they are above a certain age without repeatedly sharing their identification. The fourth uses behavioural age inference, where software attempts to predict age from activity patterns.

These methods should not be treated as interchangeable. Facial age estimation can be uncertain, especially close to a legal age threshold. Document verification may exclude people without suitable identification. Behavioural inference can make incorrect predictions, and digital credentials vary in how much information they disclose.
In October 2026, Ofcom identified photo ID matching, digital identity services, banking-based checks, and facial age estimation as methods capable of providing highly effective age assurance at 16 when appropriately implemented. It also acknowledged privacy considerations and the need for alternative checks where necessary (3).
The significant point for Canadian families is that there are multiple technical approaches. A law requiring age assurance does not, by itself, establish that extensive behavioural surveillance is necessary.
The companies developing these technologies
There is already a commercial industry developing technologies that can verify identity, estimate age, assess behavioural patterns, and detect suspicious account activity.
It is important to distinguish companies developing age assurance tools from companies specializing in behavioural fraud detection. Their products may overlap in some technical capabilities, but that does not mean every company listed below is supplying age-verification systems for social media restrictions.
These companies are examples of a broader marketplace. Some specialize in determining whether someone meets an age requirement. Others focus on whether a person using an account is likely to be its legitimate owner.
The existence of these technologies is well documented. What has not been established is that Canadian age gating legislation requires the companies involved to create comprehensive behavioural profiles of children, and that distinction is essential when evaluating the evidence.
Are we creating a new surveillance economy?
The phrase surveillance economy describes a business environment in which collecting, analyzing, and deriving value from information about people's activities becomes an important commercial function. Historically, much of the public discussion has focused on advertising. Social media companies collect information about interests, interactions, and online behaviour to personalize content and advertising. Identity and security services introduce a different commercial purpose, such as collecting or processing information to establish trust, prevent fraud, and control access.
Those purposes are not identical. Fraud prevention can be beneficial, and a company can provide identity verification without selling personal data or building advertising profiles. What is important for parents and caregiver to understand is that age-verification legislation could increase demand for identity and age assurance services. Companies may earn revenue by providing checks to platforms that must comply with new laws.
The broader behavioural biometrics market is also expanding. Grand View Research estimated the global market at approximately US$3 billion in 2025 and projected substantial growth through 2033 (4). These are commercial market estimates, not proof of the actual surveillance practices of individual companies. There are several ways a surveillance related business model could emerge.
1/ Platforms may need to conduct more age checks, increasing the number of people whose information is processed by identity providers.
2/ A provider that verifies age across multiple websites could potentially learn which services a person visits if the system is designed to reveal that information. This risk can be reduced through privacy preserving credentials and technical separation.
3/ Providers may have commercial incentives to expand their services beyond age checks into fraud prevention, account security, identity management, or behavioural risk analysis.
4/ Systems that initially collect limited information could be expanded to include additional data sources if organizations believe this improves accuracy or reduces circumvention.
These are plausible risks, not inevitable outcomes. The concern is not simply that companies will profit from age verification, providing a legitimate service for a fee is not inherently problematic. The concern is whether the commercial value of collecting more information could gradually outweigh the original purpose of collecting only what is necessary to establish someone's age. That is where legislation, independent auditing, privacy law, and technical safeguards become critically important and this has not been made clear in Bill C-34
The Canadian privacy implications are already being recognized
One of the most important developments in this debate is that Canada's own privacy regulator has identified many of the risks discussed in this article. The Office of the Privacy Commissioner of Canada has published guidance for organizations developing age assurance systems (5). It identifies several specific concerns, including personal information being compromised in a data breach, people being tracked across websites, excessive collection of personal information, and certain populations being unfairly disadvantaged by verification requirements.
The guidance also emphasizes that age assurance providers should not use information collected for age checks for unrelated purposes, subject to limited exceptions, and should avoid retaining information about the websites people visit. This is significant because it demonstrates that privacy concerns surrounding age assurance are not merely speculative objections raised by critics of social media restrictions, they are recognized by the federal office responsible for protecting Canadians' privacy.
Canada's Department of Justice has also acknowledged that Bill C-34's age verification provisions could engage rights protected by the Canadian Charter of Rights and Freedoms, including freedom of expression, equality rights, and privacy-related interests under section 8 (6). Acknowledging these potential Charter implications does not mean a court has found the proposed legislation unconstitutional, but Canadian legal scholars have predicted that such a challenge will be certain if the Bill passes with the age-gate.
To its credit, the proposed legislation includes privacy safeguards. The government's published explanation says regulated services would be required to limit the collection, use, retention, and disclosure of personal information needed for age assurance (7).
This is an important protection in principle. However, its effectiveness will depend on the regulations, technical standards, oversight, and enforcement mechanisms used in practice, for example,
what does it mean to collect only the information necessary to verify someone's age?
Does that include ongoing analysis of their behaviour?
Can a provider retain a reusable identifier?
Can information collected for an age check later be used for fraud prevention?
What happens when an AI system incorrectly identifies a 15 year old as an adult or a 17 year old as a child?
These are not minor technical details. They will help determine whether an age-verification system protects privacy or creates new privacy risks. Do we really want to pass a Canadian Bill that doesn’t first answer these question?
IMPORTANT NOTE: Dr. Michael Geist, Canada Research Chair in Internet and E-commerce Law at the University of Ottawa, points to a significant difference between what the House of Commons Standing Committee on Canadian Heritage recommended and what the federal government through Minster Marc Miller proposed in Bill C-34. The committee, which falls under Marc Miller, the federal Minister of Canadian Heritage, only recommended banning targeted advertising to children. On social media age restrictions, it called on the government to monitor policies in other jurisdictions, assess their effectiveness, identify policy options, and publish its findings. The federal government took the opposite approach, leaving out the recommended advertising ban while proposing a minimum age of 16 for accounts on certain social media services. In Geist’s view, the government’s response leaves a central question unanswered, “How did it weigh the civil liberties concerns raised by the committee when deciding to proceed with an age restriction? (8)”
The Australian experience: An important lesson for Canada
Australia provides a real world example of how social media age restrictions can operate. Its restrictions came into effect on December 10, 2025, requiring covered platforms to take reasonable steps to prevent people under 16 from holding accounts. The Australian government has emphasized that the responsibility for compliance rests with the platforms rather than children and their families. Importantly, Australia's legislation does not require every person to upload government-issued identification. Platforms must provide reasonable alternatives to government ID, and they can use existing reliable information where appropriate (9).
Australia's privacy regulator also recognizes that age estimation can involve behavioural information, including patterns of online activity and social connections. This illustrates two competing realities:
1/ Age-assurance technology can be implemented with meaningful privacy safeguards, but
2/ Some methods may still involve examining personal information that extends beyond a person's date of birth.
Australia's experience also highlights why we must distinguish between enforcing an age restriction and demonstrating that the restriction has improved children's well-being. Did the legislation establish how many individual children stopped using social media, whether they moved to other services, or whether their mental health improved, these are all important questions that need to be answered before other countries such as Canada look to implement an age gate into our legislation.
For Canadian policymakers, the lesson is that enforcement numbers, privacy outcomes, circumvention rates, and actual improvements in child safety should all be evaluated separately.
What happens when a system gets a child's age wrong?
No age estimation technology is perfectly accurate. This becomes particularly important when legislation establishes a strict threshold at age 16.
Consider two teenagers, one is 15 years and 11 months old. The other is 16 years and one month old. A facial age estimation system may have difficulty distinguishing between them. Behavioural age inference may face similar challenges because the interests and online activities of teenagers do not suddenly change on their sixteenth birthday. If a system incorrectly classifies someone as underage, that person could lose access to an account they are legally permitted to use. If the system incorrectly classifies someone as old enough, the restriction may fail to achieve its intended purpose. There is also the issue of what happens after an incorrect decision:
Will the teenager be required to upload government identification?
Provide additional photographs?
Submit information about a parent or caregiver?
Appeal to a third-party identity provider?
Each additional verification step could involve more personal information. The Office of the Privacy Commissioner specifically warns that age assurance systems may disadvantage people who lack government identification, use shared devices, or experience reduced accuracy from certain technologies. It recommends accessible and equally privacy-protective alternatives (10).
For families, the fairness of an age verification system therefore depends on more than its average accuracy. It also depends on how errors are corrected, how quickly access is restored, and whether a child must surrender additional personal information to challenge an incorrect decision.
What about adults? Why this is not just a children's privacy issue
One of the less obvious implications of age-gating legislation is that a system designed to restrict children may also affect adults. If a social media platform must determine whether a user is at least 16, it needs some reliable basis for making that decision. Sometimes the platform may already have sufficient information. For example, an account with a long, reliable history may not require an additional check. In other circumstances, adults could be asked to prove their age. This means age-assurance requirements can, and likely will, potentially extend identity checks to people who are not the intended targets of the restriction. A 45 year old Canadian might be required to complete an age check to use a service that they previously accessed without providing identification.
The privacy implications depend on how the check is performed. A system that simply confirms the person is over 16, without disclosing their identity or retaining unnecessary information, poses a different level of risk than one requiring a photograph of their passport to be stored by a third party. This is why privacy preserving age credentials deserve serious consideration. Such technologies can allow a trusted source to confirm that someone meets an age threshold without revealing their name, exact birthdate, or identification number to the website requesting confirmation.
Canada's privacy regulator encourages approaches that prevent age assurance providers from learning which websites a person visits and prevent websites from learning unnecessary details about that person. This is sometimes described as double anonymity (11). In practical terms, the goal should be to answer the question, "Is this person old enough?" without unnecessarily answering, "Who is this person, where have they been, and what else do we know about them?"
Could AI create an arms race between impersonation and surveillance?
The growth of generative AI adds another complication. AI tools can be used to produce convincing synthetic images, voices, and other forms of impersonation. Attackers may also attempt to defeat age verification systems using manipulated documents, borrowed credentials, or other techniques. As these threats develop, identity verification companies may introduce more sophisticated checks to determine whether the person presenting information is genuine. Some of these checks may involve detecting signs of image manipulation, verifying that a real person is present, examining device characteristics, or identifying unusual behaviour.
There is a legitimate cybersecurity reason for developing such capabilities. , however, there is also a risk of escalation. If attackers become better at impersonating legitimate users, providers may seek additional information to distinguish real users from fraudulent ones. Attackers may then develop new methods to evade those checks.
The concern is that this competition could encourage progressively more intrusive verification methods, but that outcome is not inevitable. Strong authentication can also be achieved through approaches such as phishing resistant passkeys, hardware security keys, cryptographic credentials, and privacy preserving verification systems.
These alternatives demonstrate that collecting more personal information is not the only way to improve security. The important question is whether companies and regulators prioritize methods that establish the necessary fact with the least possible collection of personal information.
Are we addressing the risks of social media or simply restricting access?
There is another issue that deserves attention. Age verification addresses who can access a service, it does not necessarily change the features of that service that may contribute to harm. A platform can verify that someone is 16 or older while continuing to operate the same recommendation algorithms, engagement systems, advertising practices, and content-moderation processes. Here’s a reality that we are already seeing in other countries who have implemented an age gate, a young person who is prevented from creating an account on one platform may still encounter harmful content elsewhere.
This does not mean age restrictions cannot have benefits. Supporters argue that delaying access may reduce exposure to certain risks during adolescence and give families additional time to develop digital literacy skills. The question is whether age restrictions should be considered a substitute for safer product design. It is important for us to acknowledge that Bill C-34 recognizes this distinction by proposing obligations related to child protection, harmful content, platform design, and accountability, in addition to its minimum-age provisions (12).
At The White Hatter, our approach has been to emphasize readiness-based access, digital literacy, and the role of parents and caregivers in determining when their children are prepared to use different technologies. We also believe governments have an important responsibility to regulate product design and hold companies accountable when their products create unreasonable risks.
These objectives need not be mutually exclusive. A family can support stronger protections for children while questioning whether broad age verification requirements are the most effective or privacy protective way to achieve them.
What parents and caregivers should be asking
The technical complexity of age verification can make it difficult for families to know what questions to ask. However, parents and caregivers do not need to understand machine learning or cybersecurity engineering to evaluate the basic privacy implications. Before a child or teenager is asked to complete an age check, there are several questions worth considering:
1/ What information is being collected? Is the system asking for a birthdate, a facial image, government identification, or information about online behaviour?
2/ Who receives the information? Does it remain on the device, go directly to the social media company, or get processed by a separate identity-verification provider?
3/ How long is the information retained? Is it deleted immediately after the check, or can it remain in a database?
4/ Can the information be used for other purposes? Is it restricted to age verification, or could it be combined with other information for fraud detection, profiling, or commercial purposes?
5/ Can the provider track someone across websites? Does the system know where a person is verifying their age, or can it provide proof without learning which service requested it?
6/ What happens if the system makes a mistake? Is there a simple appeal process that does not require unnecessary additional personal information?
7/ Are there less intrusive alternatives? Can someone prove they meet an age threshold without disclosing their identity?
These questions should be directed not only at technology companies but also at the governments and regulators establishing the requirements.
What safeguards should Canadians look for?
A privacy protective age assurance framework can be evaluated against several practical principles.
Data minimization: Collect only the information genuinely necessary to establish whether someone meets the required age.
Purpose limitation: Prevent personal information collected for age assurance from being repurposed for unrelated activities without a valid legal basis.
Limited retention: Delete sensitive information when it is no longer necessary rather than retaining it indefinitely.
Independent oversight: Require meaningful audits and accountability, rather than relying exclusively on assurances from the companies selling the technology.
Privacy-preserving verification: Support systems that establish age eligibility without unnecessarily disclosing identity or browsing activity.
Fairness and accessibility: Ensure that people without government identification, those using shared devices, and those incorrectly classified by automated systems have reasonable alternatives.
Transparency: Require clear public explanations of the technologies being used, the information collected, and the rights available to users.
These principles are broadly consistent with the direction outlined by Canada's Office of the Privacy Commissioner. They provide a useful basis for evaluating the eventual regulations and implementation of Bill C-34.
Let’s not forget about security concerns
When an online age check asks someone to upload a driver’s licence, passport, or selfie, it raises a question that deserves more attention in Canada’s under 16 social media debate. What happens to that information after the check is finished? If companies retain copies of identity documents, they can create a honeypot for cybercriminals. A single driver’s licence can contain a person’s full name, date of birth, address, photo, and licence number. That information can be used for identity fraud or to make a scam more convincing.
Two publicly disclosed incidents show how this risk can arise. In February 2026, Sumsub, which sells age verification and facial age estimation tools, disclosed that an attacker had gained access to an internal support system (13). The exposed information consisted mainly of names, with email addresses or phone numbers in some cases. Sumsub said its investigation found that ID images and biometric data were unaffected. In a separate incident reported in 2024, administrative login credentials for identity verification company AU10TIX were exposed (14). AU10TIX said its investigation and an external forensic review found no evidence that customer data was exposed.
The risk also extends to what happens when an age check gets it wrong. In 2025, Discord said approximately 70,000 users worldwide may have had government ID photos exposed in a breach involving its customer-support contractor (15). Those photos had been submitted to appeal age related restrictions. Even if the initial age check handles personal information carefully, the appeal process can still leave sensitive documents sitting in a support system.
A promise of strong security leaves one basic question unanswered. Why keep a copy of someone’s ID once the age check is complete?
Protecting children should not require knowing everything about them
The debate over social media age restrictions is often presented as a binary choice between protecting children and allowing unrestricted access to technology, however, that framing leaves out an important part of the conversation. How we protect children matters just as much as our intention to protect them.
Age assurance technology is not inherently surveillance, some systems can verify an age threshold while collecting very little personal information. Others may rely on more extensive analysis of identity documents, facial characteristics, online activity, or behavioural patterns.
There is an established and growing commercial market for these technologies. There are also legitimate reasons for using them, particularly when protecting people from fraud, exploitation, and harmful content. What remains uncertain is how Canada's proposed age verification requirements will be implemented, which technologies will be permitted, and whether the final safeguards will adequately prevent unnecessary collection, profiling, and tracking.
We should be extremely careful not to assume that every identity provider is building a surveillance database, just as we should be extremely careful not to assume that every age verification system is privacy protective simply because its stated purpose is child safety. Both claims require evidence.
At The White Hatter, we believe children deserve safer digital environments, parents and caregivers deserve a meaningful role in guiding their children's technology use, and Canadians of every age deserve privacy protections that are not weakened in the name of safety. The objective should not be to make children prove more about themselves than is necessary. It should be to make the digital services they encounter safer, more accountable, and more respectful of their rights. If protecting a child's privacy requires a company to collect an increasingly detailed picture of that child's identity and behaviour, we need to ask whether we are solving one problem while creating another.
The answer is not to reject every form of age assurance, it’s to insist that any system introduced in the name of protecting children must also be designed to protect their privacy, because protecting children online should never require us to surrender the very rights we are trying to preserve.
Call To Action
If after reading this article you too are concerned about the privacy and security of the addition of an age gate or 16 into Bill C-34, then send an email to your member of parliament and the Prime Minister’s Office.
Links to help you find your Member of Parliament, your member of the Senate, and the Prime Minister’s Office are provided below.
Canadian Members of Parliament: https://www.ourcommons.ca/members/en/search
Canadian Members Of Senate: https://sencanada.ca/en/senators/
Canadian Prime Minister’s Office: https://www.pm.gc.ca/en/connect/contact
Digital Food For Thought
The White Hatter
Facts Not Fear, Facts Not Fiction, Enlighten Not Frighten, Know Tech Not No Tech
References














