Understanding Social Media Bans and Bill C34’s Proposed “Pause”
Social media bans gained popularity after Australia introduced one in 2024, with the law coming into effect in December 2025, although the idea first gained momentum after publication of Jonathan Haigt’s controversial book, The Anxious Generation. According to Tech Policy Press’s Global Social Media Age Restriction Tracker, bans have been implemented in seven countries, passed in four, are being considered in 15, proposed in two and under discussion in 14 countries. The more basic idea of blocking access to certain types of content online, including copyright infringing and pornographic content, can be traced to the 1990s when the internet was commercialized.
Bill C-34 does not propose a ban, but rather a pause on social media access. It is a unique and softer approach than the blunt age-gating implemented in countries like Australia. I will examine Bill C34 in more detail below, but it is important to first understand the arguments for and against age-gating access to social media in general.
As a policy measure to protect children it provides a simple solution to an incredibly complex problem. It is action oriented, decisive and easy to explain. It is also popular, although, as Sara Grimes explains, once you dig into the data of one of the most oft cited surveys, a ban is the least popular option, with things like time limits and usage reports ranked higher. A May 2026 survey by the Centre for Media, Technology and Democracy shows strong support for a variety of measures, such as age-appropriate design (84%), liability (82%) and a regulator with the power to impose monetary penalties (72%). 76% supported bans, whether as temporary or permanent measures.
The challenge for policy makers is that it is early days for bans, Australia being the first to implement a ban 6 months ago. It is unclear whether a ban will make online spaces safer for children, and what unintended consequences such a blunt approach might catalyze. Despite this uncertainty, it is a legitimate policy option to explore, whether one favours it or not. I have sat in many rooms of late with experts strongly advocating for and against bans, and the answer I conclude, is like most things, somewhere in the middle.
The argument for social media bans is that social media is inherently dangerous for children and delaying access is the only viable option to keep children safe. There seem to be three separate justifications to this view. One is that that social media’s raison d’être is to connect people, often strangers, thus social media creates a high-risk space for criminals to contact children for grooming, luring, sextortion, trafficking, recruitment and threats connected with extremist movements, and other forms of abuse. There are countless incidents of this nature, and law enforcement and organizations like the Canadian Child Protection Centre are at the front lines battling this. The second is that the business model of social media is predatory of children. Their data is harvested to curate algorithms and advertisements designed to keep the child’s attention. The spaces are designed in “sticky” ways, with rewards, games, streaks, and tiny dopamine hits designed to manipulate users – and kids – to use the platform as much as possible. Quite simply, social media are not designed with child wellbeing in mind. The third is about childhood, and what that means. Given my age, I think of it as Goonies for Generation Alpha and younger Gen Zs. It is the belief that a delay can help give children back their childhoods, focused on real world interactions, unstructured play and self-exploration.
The data on the impacts of social media are mixed. We know there is a problem, and that some children suffer profoundly from using it, but not all children to the same extent. I think there is also a difference between harms wrought by the business models, and specific harms arising from criminals exploiting the platforms to malicious ends. Many social scientists dispute the claims in Jonathan Haigt’s book, while acknowledging a problem that needs to be solved (see Ethan Zuckerman’s work and our recent panel discussion here). Tracy Vaillancourt has undertaken important work connecting developments in the teenage brain and the unique dangers of social media (for her recent discussion of chatbots see here). We also must understand the changing dynamics of online communities, with the rise of nihilistic violent extremism (see the excellent work of Amar Amarasingam, Ghayda Hassan and David Morin for more). All of their work, and many more, provides richness to understanding what precisely is going on. In this area I am a consumer of expertise to help inform my assessment of law and policy options. In the face of uncertainty, government can act. Courts will investigate social science evidence to support a policy option but do not require it to a level of scientific certainty (see Irwin Toy v Quebec). Social media bans therefore are a legitimate policy option to address a thorny issue, and the analysis will largely turn on whether it is a narrowly tailored and proportionate response to the problem.
The arguments against social media bans rest on three primary claims. First, that bans will not work. Children are savvy and can circumvent bans, or worse, children will be incentivized to use sites and apps that are more dangerous and not covered by the ban. No ban will ever be perfect, and the question is the appropriate metrics of success. An academic study published this week reported that more than 86% of under 16s surveyed had accessed one of the banned social media platforms in the last week. Common circumvention strategies using someone else’s account, fake accounts or private browsers. 66% reported being subject to age verification measures. Does this reflect aggressively mediocre compliance by social media platforms, or a failure in the ban’s effectiveness, a sign of an improved ecosystem for youth even if imperfect (like accessing alcohol or cigarettes) or is it too early for the data to mean anything at all. It likely reflects all of those things, and I am cautiously observing developments in Australia for lessons learned for Canada. The authors’ conclusion is helpful guidance in how to digest data in the coming years:
We found limited policy implementation, efforts among adolescents to circumvent restrictions, and insufficient evidence of any substantive early effects of exposure to Australia’s Social Media Minimum Age Act 2024 on the social media use of adolescents. Lessons from previous public policy changes, however, suggest that the potential benefits of the legislative change take time to manifest and often require investment in accountability and education mechanisms to encourage compliance and uptake. The full impacts of the Act may not be evident for a decade. Nevertheless, the findings of this study provide key early insights that can guide government refinement and future actions to promote health and wellbeing.
The second argument is that bans violate children’s rights to freedom of expression. This raises questions as to the constitutionality of a ban under the Charter, which I address below. Notably the United Nations Convention on the Rights of the Child provides a blueprint that can be interpreted either way concerning a ban, as it identifies the best interests of the child as of paramount importance, including their rights to freedom of expression, freedom of thought and privacy. Social media is an important outlet for children, particularly marginalized youth, to find connection, but this group is equally the most vulnerable to harm.
Third, bans require age verification/assurance, otherwise how will social media know whether a user is a child. This certainly raises privacy and cybersecurity concerns, but debates about age verification have been stripped of nuance and reduced to a binary: if one uses an age assurance mechanism, it is an unjustifiable invasion of privacy. The reality is more complex and involves choices and trade-offs. I highly recommend The Knight-Georgetown Institute publication by Eric Rescola et al, Age Assurance Online: A Technical Assessment of Current Systems and Their Limitations, to better understand age assurance mechanisms.
It also notable that several experts who have played key roles in advancing child protection and online harms legislation have cautioned against bans, such as Jim Gamble, founding CEO of the UK’s Child Exploitation and Online Protection Centre and Sonia Livingstone, Professor at the London School of Economics. To be clear, I have spoken with many experts who also favour bans, and it is more difficult to cite data to support a ban at this stage when it is conceptually new, difficult to identify the metrics for success, and positive outcomes may not be observed for years.
From all of this, I take away the following. The most important thing we can do to protect children is make social media safer. This is the foundation of Bill C-34 and its duty to act responsibly, including special duties to protect children. A ban risks creating a false sense of safety for parents. There are no circumstances in which the entire internet will be age-gated, and we do not yet know the behavioural changes a ban will incentivize in youth, although we can expect they will start using other platforms not covered by the ban or otherwise try to circumvent the age-assurance method used. We are seeing evidence of that in Australia. The value of a ban, in my view, is that it can prompt a culture change. It will no longer be viewed as acceptable to use social media until one is 16. This may be ineffective for children already on social media, but for younger children we may see a sea change in social values. More broadly, this might force stronger safety measures by social media for children of all ages.
This is where the approach in Bill C-34 presents a unique middle path.
Bill C-34’s Social Media “Pause”
Bill C-34 would require that social media “implement adequate age-verification or age-estimation measures designed to prevent a person under the age of 16 from being able to have an account with, or be otherwise registered with, the service” (s. 27(1)). Children would therefore not be banned from watching a YouTube video, but would be banned from registering an account, similar to the approach in Australia. It operates as a pause, because the social media service could be exempt from the ban if it satisfies the new Digital Safety and Data Protection Commission that it has put in place adequate safeguards to protect children (s. 29(1)).
This flips on its head current approaches to social media regulation and bans. It states, if you want to access the market that is children, you must satisfy the regulator that your services are adequately safe for them to use. Typically regulation works ex post. The legislation may require ex ante safety by design measures, but the regulator enforces compliance after receiving reports of a failure to implement these measures effectively, followed by an investigation and corrective measures and/or a monetary penalty. Bill C-34 would require that the Commission effectively sign-off that a social media service was “adequately” safe, and only then would the ban be lifted. Of course, this assumes that social media would be interested in applying for the exemption at all. But as a regulatory approach, it provides enforcement powers that regulators often lack: the ability to enforce compliance. Fines can be viewed as the cost of doing business, and foreign-based companies might simply ignore a Canadian regulator’s corrective order.
There is a significant amount of detail to unpack here, and I will attempt to get into the weeds of some of the practical and legal challenges ahead.
The Bill would regulated large social media services and those designated as presenting a significant risk of harm (s. 6). Since the Bill would not regulate gaming or private messaging features, the ban would likely apply to large social media like Instagram, TikTok, Facebook, YouTube, X and similar. Depending on the regulations, other platforms might be captured such as Reddit. The Bill provides that social media or classes of social media to which the ban applies may be articulated in regulations (s. 28). For example, it may be that a large social media service is regulated by this Bill, but it is not subject to the ban if it is not a site predominantly used by children.
The Bill measures compliance based on implementing “adequate” age assurance measures. Australia’s legislation, in contrast, requires that social media take “reasonable steps” to prevent children under 16 from having accounts. A standard of adequacy connotes a higher standard, based on whether the measures taken are sufficient to achieve the objective of preventing under 16s from registering accounts. Australia’s approach is process oriented, focused on what would be considered reasonable steps. In practice, this could be a measurable difference in how the Commission assesses age assurance measures. The Bill identifies several criteria for adequacy, such as requiring that the approach is effective and privacy preserving, but most details are left to be coloured in by the Governor in Council through regulations setting out the criteria for adequacy safeguards, and by the Commission writing guidelines on what constitute adequate safeguards (ss. 30-31).
It is unclear what will happen if a social media service obtains an exemption and is later the subject of an investigation by the Commission and found to have failed its duty to protect children (ss. 20-21). The Bill contemplates various measures, including administrative monetary penalties (s. 83), but it is silent as to whether that service might lose its exemption and once again be banned. This presents the risk of a revolving door, where services are banned, exempt and banned, but this may be the enforcement power sought to ensure safety measures are maintained. This should be clarified in the legislation.
The Bill also does not contemplate website blocking. Let’s say a US-based company with no offices in Canada decides to ignore the ban and continue to provide its services to Canadian children under 16. If a social media service does not have an office in Canada, there are minimal ways to enforce an order under s. 81. Since it is a US-based company it may be difficult to enforce the order in a US court under s. 230 of the Communications Decency Act. I do not think enforcing a ban violates s. 230, but arguably neither does an order to block content, and the Supreme Court of Canada decision in Equustek v Google was famously held unenforceable by a lower court in California based on s. 230. Website blocking may be the only avenue to force compliance, but it is a tool of last resort given its extraordinary impact on freedom of expression. It was the subject of debate concerning Bill S-209, which contemplated website blocking for pornography sites that ignore age verification requirements. We should be alert to it here, particularly since courts have inherent jurisdiction to order website blocking. If it is not addressed in legislation, it will be by courts eventually, and courts may not appreciate the high stakes and nuanced free expression concerns that drive this analysis (it was notably absent in Teksavvy Solutions Inc v Bell Media Inc).
The “ban” or “pause” rests entirely on implementing age assurance measures. As I discussed in a previous Substack article, the criteria for age assurance should draw from Bill S-209 to strengthen the privacy and cybersecurity requirements. More fundamentally, the public should debate age assurance as a mechanism for child safety. Any effort to create age-appropriate spaces for children requires the provider knows who the kids are. Otherwise, the social media provider would need to cleanse their site of anything inappropriate for children. This is not the internet we want and therefore we should consider more carefully the risks and trade-offs of various age assurance measures.
There are various types of age assurance mechanisms, and the Bill contemplates age verification and age estimation measures. While they all necessitate data collection, this can be done in a privacy-preserving way but entails trade-offs. Age verification is the most accurate and privacy invasive, such as collecting someone’s ID. This should not be the first step to assess someone’s age. Age estimation is less accurate but more privacy preserving, such as looking at behavioural cues, and facial features. Children at the cusp of 16 will often be incorrectly assessed as younger or older than they are. Other signals can assure a provider as to age, such as how long one has had a Gmail account, Facebook account (2007 for this author) or similar. Australia uses a waterfall approach, where the least privacy-invasive measure is used first, and a more invasive approach is only used if necessary because the initial approach was inconclusive.
Arguments against age assurance tend to view the use of this technology as in and of itself an unjustifiable privacy invasion. Other arguments are specifically concerned with age gating some of the most popular websites and apps, and what this will require of adult users. These will all be important points of debate when this Bill is studied at Committee. In my view, age assurance is one of the most important policy questions in the coming years. My key concern is not whether age assurance is used – it is even now by many social media providers - but rather, that whatever method is used is the most privacy-preserving and cybersecure approach to achieve the objective. This requires getting into the weeds of what specifically is being asked of companies, and what accountability is in place to ensure they do not implement sloppy practices. The technology has advanced significantly, and sometimes inaccuracy is the best and most adequate method to preserve privacy while still identifying most children to put in place age-appropriate features or a ban. Bill C-34 identifies key privacy requirements, such as limits on data use, retention and swift destruction after used (s. 27). It does not go far enough and should require reasonable security safeguards and that assurance is performed by a third party independent of the social media service.
Is the “Pause” Proposed in Bill C-34 Constitutional?
In the lead up to the Bill being tabled, I had the opportunity to meet with free expression scholars from across Canada and, among other things, we discussed the constitutionality of bans. There were mixed views. I recommend reading Professor Emmett McFarlane’s analysis. While I agree with much of his analysis, I do not reach the conclusion that the ban is clearly unconstitutional. In my view, a court could go either way on this, but the “pause” as framed is likely constitutional.
A law requiring that a social media service restrict access to children without a doubt violates s. 2(b) of the Charter. The right to freedom of expression includes the right to seek, receive and impart information and ideas, and the law would restrict children’s ability to do so. It does not matter that this is a commercial site, but rather that the government is restricting access to the platform of their choice for expressive purposes.
The main analysis is whether the interference can be justified under s. 1. This requires examination of whether the objective of the law is pressing and substantial, whether the approach taken can be rationally connected to this objective, whether the right is limited as little as possible, and whether as a whole, the benefits outweigh the risks. The best case for reference is Irwin Toy v Quebec, where the SCC upheld a ban on television advertising to children, but it does not wholly translate to social media, as TV advertising pushes information to viewers, while social media enables the spectrum of expression, from consuming to sharing and creating. There are a few lessons from that case. First, the Court was deferential to policy decisions by government, and while requiring evidence to justify the approach taken, did not require it to a scientific standard. The Court also took seriously the vulnerability of children to manipulation from advertising. However, the dissent was formidable and was unpersuaded by the evidential basis for the ban.
The “pause” approach makes this less susceptible to constitutional challenge. The objective of child safety is pressing, and requiring that companies implement safety measures is defensible and connected to achieving that goal. The riskier questions are as to minimal impairment and proportionality. Courts do not require that the approach taken is the least minimally impairing choice, but rather that it is reasonable among a range of options. The Bill already proposes a duty to protect children and to implement safety by design, and an argument might be made that in light of this, a ban is unnecessary and overreaches. However, I expect that a court will likely examine closely the risks of social media use for children and view the exemption as a defensible policy approach within the range of minimal impairment, even if it is not the least impairing.
In terms of proportionality, this is where a court would consider the various factors explored in this article about the benefits and negative impacts of the ban. A court would consider whether a ban would be effective, and the privacy impacts of age assurance on both children and adults, and the overall impact on freedom of expression of children. A court may weigh heavily the question of whether this ban will be effective given its narrow focus on a small list of social media services and ease of circumventing bans. The question is how this might be balanced against the age assurance mechanisms that companies will be required to implement. These mechanisms certainly implicate privacy, but that is only the beginning of the analysis, and privacy is a fundamental right but is not specifically protected under the Charter s. 8 outside the context of search and seizure. Weighing the different age assurance options available to social media, and the privacy requirements in the legislation, I do not think privacy will be a barrier to the constitutionality of the social media restriction, in particular because of the availability of an exemption. If anything catches the courts attention, I expect it may be the effectiveness of these bans balanced against the broad array of potential negative impacts.