Why Bill C-34’s Social Media Ban for Kids Is Unconstitutional

In my last episode about Bill C-34, the SafeSocial Media Act, I said that one provision jumped out at me as particularly constitutionally vulnerable: the rule that would keep people under sixteen from having accounts on social media services.

I said then that social media is fundamentally expressive. Teenagers have freedom of expression too. And telling a fifteen-year-old that the government will not allow them to have an Instagram account is very different from telling them they cannot buy a bottle of whisky.

Well, since then, something very interesting has happened in France.

France passed a remarkably similar law restricting social media use by young people. And on August 14, 2026, the French Constitutional Council struck down the core of that law as unconstitutional.

Canada is not France. The French Constitution is not the Canadian Charter of Rights and Freedoms. And the legal tests are not identical. But when you read the French decision, the reasoning sounds remarkably familiar to a Canadian constitutional lawyer.

Are there legitimate issues with social media, particularly for kids? Yes.

Is addressing this a “legitimate government objective”? Yes.

But a broad prohibition on young people using social media?

It is too blunt an instrument. It is too broad. And it is not sufficiently connected to the actual risks.

In Canada, I think we would reach much the same result under section 2(b) of the Charter and, particularly, the minimal impairment part of the Oakes test.

And today we’re going to do some constitutional law. Yippee!

Part One: What Bill C-34 Actually Does

First, let's be precise about what Bill C-34 says.

It does not literally say that every person under sixteen is forbidden from ever looking at social media.

Instead, section 27 says that operators of social media services selected by regulation have to use age verification or age estimation measures designed to prevent anyone under sixteen from having an account or otherwise being registered with the service. Section 28 allows the government to prescribe additional measures to accomplish exactly the same thing.

That distinction matters.

A fifteen-year-old might still be able to look at some publicly available material without an account.

But they cannot post. They cannot comment. They cannot participate through an account. They cannot use the platform in the way that social media is fundamentally designed to be used.

And there is an “escape hatch” that makes this less absolute.

Under section 29, the operator of a social media service can apply to the Digital Safety Commission for an exemption from the ban if the Commission concludes that the service provides “adequate safeguards” for children. The government can make regulations establishing the criteria, and the Commission can issue guidelines.

But notice what this exemption is not.

It is not an exemption for a particular fifteen-year-old.

It is not parental consent. (Parents have no say over any of this.)

It is not an assessment of whether this particular young person can safely use this particular service.

It is essentially a service-by-service exemption. The platform gets exempted, not the child.

Part Two: Section 2(b) of the Charter

The first Charter question is relatively straightforward.

Section 2(b) protects freedom of thought, belief, opinion and expression. It also includes freedom of “other media of communication.”

Fundamental freedoms

2 Everyone has the following fundamental freedoms:

(a) freedom of conscience and religion;

(b) freedom of thought, belief, opinion and expression, including freedom of the press and other media of communication;

(c) freedom of peaceful assembly; and

(d) freedom of association.

The Supreme Court of Canada has interpreted “expression” extremely broadly. The basic rule coming out of cases like IrwinToy is that an activity is expressive if it attempts to convey meaning.

Posting a video is expression. Writing a comment is expression. Participating in a political discussion is expression. Sharing your artwork is expression. Talking about your school, your community or your favourite hockey team is expression. And social media has become one of the principal places where people do all of those things.

If you just look at the definitions of “social media service” in the Bill, you can see that what makes a service “in scope” of the Bill is the fact that it enables accessing, sharing and communicating content. That is expression.

In fact, Bill C-34 itself recognizes the importance of freedom of expression but perhaps in more of a “window dressing” kind of way. One of the statutory purposes is to mitigate online harms while respecting freedom of expression, and another is to enable people in Canada to participate fully in public discourse and exercise their freedom of expression online.

The Bill even tells the Digital Safety Commission that when it makes regulations and guidelines, it has to take freedom of expression into account.

The social media ban for kids doesn’t take freedom of expression into account.

I don't think the government gets very far arguing that section 2(b) isn't engaged. It is clearly engaged.

The government is imposing a legal rule that prevents a class of people from obtaining the accounts they need to participate in major forums for expression. That is a restriction on expression, plain and simple.

Which takes us to section 1 of the Charter.

1 The Canadian Charter of Rights and Freedoms guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.

Part Three: The Oakes Test

A violation of a Charter right is not necessarily unconstitutional.

Section 1 allows governments to impose reasonable limits on Charter rights if those limits can be demonstrably justified in a free and democratic society.

The test to determine if a government measure that limits a Charter can be demonstrably justified in a free and democratic society is called the Oakes test.

Very simplified, the government needs an important objective. The measure chosen by the government has to be rationally connected to that objective. It must impair the Charter right no more than is reasonably necessary. And, finally, the benefits have to justify the harm to the protected right.

I don't think the government is going to have much trouble at the first stage.

Bill C-34 expressly says one of its purposes is protecting children's physical and mental health.

There are legitimate concerns about young people and social media. There are legitimate concerns about harmful content, addictive or manipulative features, bullying and other risks. I am quite prepared to assume in this analysis that protecting children from those harms is a pressing and substantial objective.

I would probably give the government “rational connection” as well. If you keep children completely off a service, they obviously cannot be harmed by that service while they are off it. But that is far from the end of the analysis.

The really difficult question is: Do you have to keep them off the service altogether?

That's minimal impairment. And that is where I think this provision is in serious trouble.

Part Four: Minimal Impairment Does Not Mean “Anything That Works”

The Supreme Court does give governments some latitude here.

Minimal impairment does not mean that a judge gets to dream up some theoretically perfect alternative and strike down legislation because Parliament did not choose it. Governments are generally allowed a range of reasonable policy choices. But there is still a constitutional limit.

The question is whether the government can achieve its objective, in a real and substantial way, through a measure that interferes significantly less with the Charter right.

The Supreme Court's decision in RJR-MacDonald is particularly interesting. That case dealt with a broad prohibition on tobacco advertising. The majority emphasized that a complete prohibition on expression is more difficult to justify than a partial prohibition, particularly where the government has not demonstrated why narrower measures would be inadequate.

And in a case called Carter, the Court made essentially the same point in a very different context: an absolute prohibition was not minimally impairing where a carefully designed system of safeguards could substantially accomplish the government's protective objective. 

Now, the government has a very good case it will undoubtedly rely upon: Irwin Toy. That was the famous case involving Quebec's prohibition on commercial advertising directed at children under thirteen. The Supreme Court found a section 2(b) violation but upheld the restriction under section 1. 

But there is an enormous difference.

Irwin Toy restricted what advertisers could say to young children. It regulated commercial expression directed at a particularly vulnerable audience. Bill C-34 restricts what young people themselves can do.

It removes their ability to participate through an account in an entire medium of communication, including political, educational, artistic, cultural and ordinary social expression.

Those are very different Charter interests.

Part Five: And Then There Is France

This is where the French decision becomes really interesting.

First, a quick note about the French Constitutional Council: its mandate includes reviewing certain laws after they are passed by the legislature but before they are signed into law. It’s kind of like an automatic reference to the Supreme Court. I think that’s a cool protection for constitutional rights.

Back to the social media ban: The French Constitutional Council began by recognizing essentially the same thing that I was just talking about: online communication services have become important to participation in democratic life and the expression of ideas and opinions. It concluded that freedom of expression therefore includes the freedom to access and express oneself through those services.

Then it acknowledged the government's legitimate concern.

The legislature was trying to protect young people from risks associated with social networks. The Constitutional Council accepted that protecting children could justify some restrictions.

But then it looked at what the law actually did.

The prohibition was not sufficiently tied to the particular functions of a service, the content it offered, the dangers it presented or the adequacy of its protections. It could therefore apply to services where a risk to children's health or safety had not actually been established. And the law treated all children under fifteen alike.

There was no mechanism for a parent, properly informed about the risks and safeguards, to authorize access. There was no individualized consideration of the child's age, maturity or circumstances, or of the particular service involved.

The Constitutional Council therefore concluded that the general prohibition was not appropriate, necessary and proportionate to the objective.

And it struck it down.

Does that decide the Canadian question? Nope. But it is almost a textbook description of the minimal-impairment problem with Bill C-34’s social media ban.

Part Six: The Canadian Law Is Different — But Not Different Enough

There are some important differences between the French law and Canada’s Bill C-34..

The Canadian scheme is somewhat more targeted. The government has to specify the regulated social media services or classes of services to which the under-sixteen rule will apply. This is solely based on the number of users, so has nothing to do with specific risks to specific kids.

And Canada has the section 29 exemption for services with adequate safeguards. (Which will likely only come into effect years after the bill becomes law and the Digital Safety Commission gets its stuff together.)

Bill C-34 also deals with the privacy implications of age verification more explicitly than the French law did. It limits the purposes for which age-verification information can be collected and requires its destruction after verification or estimation is completed.

Those differences matter, but they don’t solve the central section 2(b) problem.

Once a service has been selected, the rule right out of the gate is still: if you are fifteen, you don't get an account.

It doesn't matter why you want the account. It doesn't matter whether you're using it to talk about politics, follow community organizations, publish your photography, participate in a school activity or just talk to your friends.

It doesn't matter whether your parents think you're capable of using it responsibly.

And it doesn't matter whether the particular feature you want to use has anything to do with the harm that Parliament is trying to prevent.

That is a very broad response to a much more specific collection of problems.

Part Seven: The Strangest Part Is That Bill C-34 Already Contains the Alternatives

And this is what I think makes the minimal-impairment argument especially powerful.

Bill C-34 itself contains a whole range of less intrusive tools for protecting children.

Section 21 allows age-appropriate design requirements. Section 32 requires social media operators to implement measures to mitigate exposure to harmful content. Other provisions require blocking tools. And the Commission's regulation-making powers expressly contemplate things like children's account options, parental controls and other age-appropriate design features.

In other words, Parliament has already written a menu of more targeted interventions into the same statute.

• Protective defaults.
• Parental controls.
• Age-appropriate account settings.
• Restrictions aimed at particular risky features.
• Measures aimed at harmful content.

Those approaches regulate the risk.

The under-sixteen ban regulates the person.

And that is exactly the question the government is going to have to answer in a Charter challenge: why aren't those less rights-infringing measures enough?

The government does not have to prove that no conceivable alternative exists. But it does have to justify the level of impairment it has chosen. And saying “social media can be harmful to kids” doesn't answer that question.

Conclusion

None of this means that social media is harmless.

And it doesn't mean that Parliament is powerless to regulate it.

The Charter does not require the government to sit on its hands while children are exposed to risks. It does, however, require the government to respond to those risks with some degree of precision.

Target the dangerous features.

Target the harmful content.

Require safer design.

Give parents useful controls.

Impose obligations on the platforms that are actually creating the risk.

But telling every fifteen-year-old that they cannot have an account on a designated social media service is something very different.

France's Constitutional Council has now confronted essentially the same constitutional tension. It accepted that the harms were real. It accepted that protecting children was legitimate. But it concluded that those legitimate concerns did not justify a general prohibition that swept far beyond the risks the government was trying to address.

Canadian courts will have to take a different constitutional route, but I am pretty confident they will arrive at the same destination.

If Bill C-34 is enacted in its current form and the under-sixteen account restriction is challenged, I think minimal impairment is where it breaks.

And I don't think putting an exemption process at the end of an overly broad prohibition fixes the problem at the beginning.

I realize that a social media ban for young people is very popular and politically wonderful, but we exist in a country with rights. A fifteen year old has the same right to freedom of expression as I do. I can have an account on YouTube to share my thoughts on privacy, technology and important public policy matters. Once Bill C-34 is passed – if it is passed in its current form – young people will not be able to do that. And that, in my view, is a serious issue that needs to be front and centre in the discussion of Bill C-34 when Parliament gets back from its summer break.

And I should note, as I wrap up, that I also think this violates adults’ Charter rights, because adults will have to jump through privacy-risking age verification measures before being allowed to use social media. I’m not sure that can be fully justified under the Oakes test, either.