Beyond the Ban: Can We Protect Children on Social Media Without Denying Them Access?
29 octobre 2026
Oversight Board Position Paper
I. Introduction
Protecting children* online is one of the most pressing policy challenges facing the tech industry today – and arguably society more broadly. Lawmakers, whistleblowers, tech companies, parents, educators and a wide range of advocacy groups have over years raised serious concerns about the dangers social media presents to children, in particular between the ages of 13-17. The 2026 landmark settlement between Meta and 52 attorneys general across U.S. states, territories and the District of Columbia, signals growing pressure on companies to address design choices in their products that many see as harmful to children. Now, the question is twofold: first, whether platforms’ design choices effectively address the serious risks children face online in all parts of the world and, second, what this means for social media bans that claim to seek more protection for young people but may end up denying them the benefits these tools can offer.
* In the context of this discussion, the Board is specifically focused on children between the ages of 13-17. While the Board recognizes the Convention on the Rights of the Child’s definition of a child as those under the age of 18 and understands that social media bans impact young people of all ages, we believe that those in this age range face some of the most specific challenges with these laws.
Lawmakers, whistleblowers, tech companies, parents, educators and a wide range of advocacy groups have over years raised serious concerns about the dangers social media presents to children.
Most major social media companies, including Meta, have determined that children under 13-years-old should not have access to their platforms (though research suggests that many are accessing them nonetheless). Parents have been offered technological solutions by social media companies, including a range of parental controls and time limits, to help manage their children’s platform access. In recent years, governments have considered – and in several cases enacted – legislation seeking to bar young people from social media entirely, beginning with Australia’s Online Safety Amendment Act of 2024. The legislation bars users under 16 from social media platforms such as Instagram, Facebook, Snapchat, TikTok, and more.
Freedom of expression and access to information can play an essential role in children’s development and enjoyment of their rights, including on social media. However, social media can also pose real risks to safety, where different rights can be in tension. Navigating the tensions between freedom of expression and other rights is central to the Oversight Board's work. The Board recognizes that while the urgent concern to protect young people from the harms of social media is well placed, this protection cannot come at the expense of other children’s rights. Blanket bans on access to social media are too blunt an instrument to protect children.
In this paper, we apply the international human rights framework to social media bans, which raises several questions and concerns about their impact on children’s rights. We support an approach that expects more from companies and policymakers to protect children’s rights online while addressing real safety concerns.
II. Background
In December 2025, Australia’s law went into effect, making it the first country to ban social media for children under 16-years-old. In anticipation of the ban, Meta removed over half a million accounts of young users. Since then, several countries have followed Australia’s lead: the UK announced a social media ban to go into effect in 2027; Malaysia a comparable ban was due to go into effect in June 2026 (but has been postponed until January 2027); and France enacted a law in 2026 banning social media for young people under 15-years-old that was subsequently ruled unconstitutional on freedom of expression grounds. Denmark, Spain, Greece, and others are all projected to pass their own social media bans. Several states in the U.S. – for example, Florida – have also passed comparable legislation.
Polling in March 2026 showed that three in five Australians believe the ban to be effective thus far.
Months into Australia’s ban, the effectiveness of the legislation is still unclear, with young people finding workarounds to maintain their access to platforms. Australia’s eSafety Commission, the national independent government regulator for online safety, reported that seven in 10 parents whose children already had a Facebook, Instagram, Snapchat or TikTok account before the ban came into effect said their teens by February 2026 still had access to at least one of the age-restricted services. While the tangible impact of these laws on young people is still being determined, public demand for action remains high and the popularity of bans continues to rise among lawmakers and the public. In response to social media companies’ failures to demonstrate their products and content policies safeguard children’s rights, these blanket bans have become a popular default – appearing to offer a straightforward solution in the absence of clear alternatives. Polling in March 2026 showed that three in five Australians believe the ban to be effective thus far. Research published in July 2026 showed that a majority of Americans support a social media ban.
III. A Rights-Based Approach
While keeping young people safe online is a fundamental concern, any restrictions should be considered in accordance with international human rights principles. Both the benefits and harms of removing young people’s social media use need to be deliberated in parallel. Social media bans present a binary choice between access or no access, but the Board argues that the most serious harms can, and should, be mitigated in other ways, with less cost to freedom of expression for young people.
When restrictions on freedom of expression are at stake, international human rights law provides a test to assess the legitimacy of those restrictions.
Three-Part Test to Assess Restrictions on Freedom of Expression
When restrictions on freedom of expression are at stake, international human rights law provides a test to assess the legitimacy of those restrictions. This test looks at three dimensions: legality, legitimate aim, and necessity and proportionality of any potential restriction to freedom of expression. It is reflected in Article 19 of the International Covenant on Civil and Political Rights (ICCPR), and in Article 13 of the UN Convention on the Rights of the Child (CRC). Many responses to the risks posed by social media to children carry potential consequences for their rights, including those essential to safety – the human rights framework provides guidance on how to resolve tensions between rights. Using the three-part test to assess social media bans for their impact on children’s freedom of expression reveals some of the Board’s key concerns that such bans are ultimately overbroad.
Legality requires rules restricting expression to be clear and accessible to those they apply to and to those implementing them. Early legal challenges to the Australia ban, such as Reddit’s, have contested whether their services meet the definition of a social media platform, showing the kinds of uncertainty these prohibitions can produce. For this reason, analyzing each legislative proposal to ensure it meets the legality test is essential, although the Board’s main concerns about social media bans are not focused on the clarity or accessibility of such measures.
Legitimate aim requires any restriction on freedom of expression to also pursue one or more of the legitimate aims listed in Article 19 of the ICCPR and Article 13 of the CRC, including to protect the rights of others. These laws, to the extent they are aimed at protecting the rights of children essential to their safety and well-being, can be pursuing a legitimate aim.
The bans - which are intrusive by design - have been applied over several less intrusive means, including safety-by-design approaches by platforms or better content moderation efforts by companies.
The principle of necessity and proportionality, as defined in ICCPR General Comment No. 34, requires that restrictions on expression “must be appropriate to achieve their protective function; they must be the least intrusive instrument among those that might achieve their protective function; they must be proportionate to the interest to be protected.”
It is here that the discussion around social media bans needs to be most nuanced. While protecting children from harm is a legitimate aim, that does not justify all restrictions. The bans – which are intrusive restrictions by design – have been applied over several other less intrusive means, including safety-by-design approaches by platforms or better content moderation efforts by companies. This may include targeted age restrictions on particular product features and functionalities, as well as age-specific content restrictions, provided all parts of the three-part test are satisfied.
The principles of the “best interests of the child” (Article 3 of the CRC) and the “evolving capacities” of children (Article 5 of the CRC) must also be taken into account as part of the necessity and proportionality assessment. Regarding children’s evolving capacities, Article 5 of the CRC recognizes that children’s abilities to exercise their rights develops progressively with age and maturity. General Comment No.25 extends this principle to the digital environment, stating that, “states parties should respect the evolving capacities of the child as an enabling principle that addresses the process of their gradual acquisition of competencies, understanding and agency. That process has particular significance in the digital environment.” This principle adds another layer of complexity to a necessity and proportionality analysis. What might be considered the “least intrusive” measure may be different for children of different ages and in different contexts, and should consider myriad factors relevant to a child’s development and maturity.
Even if the bans reduce harms in some respects, the Board's key concern remains: do these benefits outweigh the cost of bluntly cutting off an entire group from all social media platforms?
Key Concerns
Assessing social media bans as currently designed and implemented using frameworks such as the necessity and proportionality test, reveals procedural and substantive concerns about whether they adequately protect the rights and safety of young social media users.
Children are rights holders
International human rights law including the CRC[1] recognizes young people as rights holders, not just objects of protection. They have the right to access information and freedom of expression, all in relation to their evolving capacities. The UN Human Rights Office and UNICEF have raised concerns that blanket age-based prohibitions on access to social media can undermine the human rights of children unless narrowly tailored to address clear harms that result from particular services or features. Child safety is critical but must be addressed holistically by protecting the fundamental rights of all children, including freedom of expression and access to information. This requires ongoing child-focused human rights due diligence (or risk assessment) to identify and mitigate harms, as well as access to remedy for those harms where they occur.
Bans remove incentives for companies and policymakers to build safer platforms
Blanket bans lessen the pressure on platforms and regulators to build less restrictive and more effective safety measures that protect children’s rights globally. Insisting on accountability for platform design is critical to developing platforms that protect children’s rights and therefore their safety. Solutions that focus on keeping children off platforms remove the company’s responsibility to make the online environment safer. They reduce pressures on social media platforms to design and operate platforms that are safe and respect children’s rights. Children who live in jurisdictions where bans are not in force may also be exposed to platforms that have not been designed with their rights in mind.
Bans remove parents from the decision-making process
Human rights law affirms not only the rights of children but also that “the family is the natural and fundamental group unit of society,” according to the Universal Declaration of Human Rights. Respect for the evolving capacity of the child also requires parental involvement, according to CRC Article 5. Board Members agree that parents need more support to navigate potential risks for their children online, especially considering the severity of potential harms on social media. Even with support, the challenges often outweigh the sphere of parental control. Parents cannot always navigate all the dynamics at play – whether due to gaps in digital literacy, competing demands on their own time or other barriers to engage – which means that guardrails need to exist beyond the family unit.
Bans on social media can easily be evaded
It is unclear the extent to which banning children from social media effectively limits their access, and whether doing so makes them safer. Since bans were only enacted starting in December 2025, research on their effectiveness and long-term impact is nascent and in some contexts nonexistent. The non-governmental organizations, Family Online Safety Institute (FOSI) and the Molly Rose Foundation, both published research showing that in Australia, despite the ban, young people still found ways to access social media accounts. FOSI’s research showed that 50% of the under-16-year-olds they interviewed were still accessing at least some of their social media accounts; 17% still had access to all their accounts. Even with blanket bans, platforms need stronger guardrails to handle circumvention. The Cato Institute has also cautioned that when we push young people off mainstream and regulated platforms, we run the risk of them moving to smaller, less moderated spaces with fewer safety features. If that is the case, there is a risk that these bans will conversely make it harder to protect children online.
Another important consideration is how policies built in more democratic countries impact human rights when copied by more repressive regimes. Australia’s ban requires a successive validation model – an ongoing verification process which starts with age inference signals and then moves to document verification or facial age estimation, if needed. The enforcement of these sorts of age restrictions or bans requires verifying age at scale and continuously, which then requires the development of identity infrastructure. Even in a rights-respecting country with strong data protection laws and practices, and an independent judiciary, this represents a privacy risk for all users. In a country with a surveillance state, weak data protection and a history of using digital tools against journalists, activists and women, that same infrastructure could become a serious threat. In any context, a mandate to know the real identity behind every account may aim to protect children, but it also requires the tooling for surveillance of everyone. Anonymity isn't a loophole to be closed; it's often a critical safety feature.
We must also consider the impact restricting access for young people can have on different societies and communities. Young people’s rights to freedom of expression and political participation could be seriously limited. And if bans are embraced in democracies to protect children, autocrats may seize the opportunity to introduce similar bans in their countries, not to protect the youth, but to protect their own grip on power.
Bans negate important forms of safety and support for some children
These laws do not take into consideration the forms of safety and support that may exist for children through their online experiences, including through how they express themselves, how they learn, play, form connections with others and access information on important topics. This demonstrates the importance of a rights-based approach not pitting access to social media or freedom of expression squarely against safety and related rights. Some children may be safer off social media, but there are also young people who may be in more peril without access to online communities, such as isolated LGBT+ youth, youth living with a disability or those in rural settings where accessing both information and social connection is difficult. During COVID-19 lockdowns around the world, social media provided almost the only opportunity for many teens to have social interaction. Additionally, the issue of too much connectivity is not a global one. Two-thirds of school-age children worldwide still lack reliable internet access. For many, connectivity is fragile and limited. When talking about banning children from digital spaces as if access were a luxury, a reality is described that does not match most children's lives. For those who can access the internet, these platforms often offer educational resources and opportunities that would otherwise be inaccessible for people around the world. Research from the National Academies of Sciences, Engineering and Medicine shows that social media should not be considered uniformly positive or negative for adolescent health – there are benefits such as social connection and identity development. It is limiting to treat the internet solely as a danger that can be switched off for some.
IV. Recommendations
The preceding analysis finds that blanket bans on children’s access to social media, however well-intentioned, are ill-suited tools for the harms they aim to address. The analysis does not suggest that all government intervention is unwarranted but rather that blanket bans misidentify how harms on social media should be addressed. With these shortcomings in mind, the Board offers the following recommendations to policymakers considering bans and regulatory requirements, and to companies designing the platforms in question.
- Insist on age-appropriate platform safety by design. The design of platforms is crucial for the well-being and safety of children. As the Meta settlement referenced above indicates, targeted and age-appropriate platform design is critical to a safer online environment for children. As part of the settlement, Meta has agreed to implement new platform design features such as time limits and nighttime blocks. Platforms should continuously improve their platform design to ensure that the correct guardrails – including time limits, notification blocks and more – are in place for children. Companies must also continue to invest in platform design that protects the rights and safety of young people and their evolving capacities.
- Involve children in decisions that affect them. Consistent with Article 12 of the CRC, policymakers and companies should have ongoing consultation with young people into the development of legislation and of platform safety features.
- Aim for more accurate content moderation: Since its inception, the Board has advocated for more accurate content moderation that respects the rights of users. This has become more urgent in the era of AI-driven content moderation. As the Board has previously said, automated moderation and curation systems must be rigorously and continually evaluated on their performance. This is particularly important to ensure that children are not exploited or exposed to harmful content on platforms.
- Invest in parental controls: Develop these tools intentionally and provide more resources to train parents and caregivers on how to use them. Caregivers need help in safeguarding and educating their children, especially in the face of potentially harmful social media design and content. Policies to address those harms should consider empowering families – giving them this important tool, along with the information and education they need to foster young people’s evolving capacities – rather than blanket bans.
- Tighten data protections Research shows that parents often overestimate the amount of data protection provided by platforms. Platforms and regulators need to create stronger guardrails for data protection and ensure the protection of the right to privacy for young people. Children and their parents need more transparency and clear communication to users on how data collection works, especially as many of the proposed measures (including bans) require sharing identifying information for age verification.
- Meaningfully address noncompliance: Regulators should ensure that companies and platforms failing to meet these standards face appropriate consequences. This could include enforcing more targeted age restrictions on noncompliant platforms, which may help drive children towards platforms that meet these standards.
- Invest in additional research: As mentioned above, there is currently limited research on the effects and efficacy of social media bans as compared to other measures to reduce harm. To best protect young people on platforms, policymakers and companies alike need to invest in more research to understand the effectiveness, consequences and impact of bans and other safety by design measures on young people and their families. As more bans go into effect, their impact should be tracked and analyzed. This research should take into consideration the varied experiences and circumstances that young people live in around the globe.
V. Conclusion
While blanket bans on children’s access to social media respond to genuine calls for much needed action, they are blunt and overbroad instruments that do not consider the evolving capacities of children, and could harm children’s rights and safety more than they advance them. A human-rights-based approach requires policymakers and companies to make more nuanced decisions that protect children’s rights as the surest way to protection and safety.
About the Oversight Board
The Oversight Board is an independent body of experts that holds Meta accountable for protecting freedom of expression and other human rights on its platforms. The Board makes binding decisions on individual content and provides policy recommendations to which Meta is obligated to respond. It is currently comprised of 20 experts from around the world, from a variety of political, cultural, and professional backgrounds.
Research Note
The structure created for the Board is designed to protect the independence of Board Members and to allow them to make judgments free from influence or interference from Meta. The Board is funded via an irrevocable trust, managed by trustees who are separate from Meta. Funds from Meta are placed into the trust and the trustees are responsible for making decisions about how the funds are spent. This trust is not under the control of Meta. Meta had no role in this position paper.
[1] Not ratified by the U.S. but in force in most of the world.
