A split that is less clean than advertised
Canada is dividing hate-speech law in two. Criminal punishment of hate propaganda now sits in Bill C-9, the Combatting Hate Act. Platform duties are meant to sit in Bill C-34, the Safe Social Media Act. The split is sensible in principle. It is not as clean as it has been described.
Per LEGISinfo, the House of Commons gave C-34 second reading on 29 October 2026 and referred it to the Standing Committee on Canadian Heritage. The Hub's preview of that day's debate said passage was not in doubt given the government's majority. It said the live argument was over the age threshold and the definition of a regulated platform.
The hatred question has a plain answer. The bill text defines "harmful content" as seven categories. One of them is "content that foments hatred." Another is "content that incites violence." The commentator who flagged this was right.
The case for the government's approach
The strongest argument for C-34 is that platforms shape what spreads, and criminal law is a blunt instrument for that. A prosecutor can reach a speaker after the fact. A systemic duty can push a platform to change its design before harm accumulates. Child-safety advocates also argue that voluntary moderation has failed. Giving a regulator, the Digital Safety Commission of Canada, the power to require risk management is a defensible answer to that.
C-34 is also a retreat from its predecessor. The Online Harms Act (C-63) died in 2025. Michael Geist's analysis of C-34 notes that the maximum administrative penalty falls from the greater of 6% of global revenue or $10 million to the greater of 3% or $10 million. The bill also precludes imprisonment for offences (clause 111). It does not reuse C-63's most controversial feature. Those are real concessions.
What C-9 actually does
The Department of Justice describes C-9 as creating offences for intimidating or obstructing access to places of worship, schools, community centres and similar places. It makes hate-motivated crime a specific offence and defines "hatred." It adds a hate propaganda offence for wilfully promoting hatred against an identifiable group by displaying certain terrorism or hate symbols in public. It also repeals the good-faith religious opinion defence. It received royal assent on 18 June 2026, and the provisions took effect on 18 July 2026.
That is a criminal regime with a mental-state requirement (wilful promotion), a prosecutorial gatekeeper and a court. This is the traditional place to draw the line on speech. A person must act deliberately, and a judge tests the allegation against the Charter.
Where the seam opens
The two bills use different tests for related conduct. C-34 defines hatred as content that "expresses detestation or vilification" on a prohibited ground under the Canadian Human Rights Act. Section 2(3) says content does not qualify "solely because it expresses disdain or dislike or it discredits, humiliates, hurts or offends." That is a sensible carve-out, and it echoes the Supreme Court's narrow reading of hatred in earlier cases.
But a platform is not a court. It must apply that definition at scale, under a regulator's oversight, with a fine behind it. A criminal statute asks whether a speaker wilfully promoted hatred. A platform duty asks whether a service managed the risk that such content would circulate. The second question has no intent element and no judge. Companies facing penalties of up to 3% of global revenue will rationally over-remove borderline political speech rather than defend it.
There is also a scope problem. Geist reports that the under-16 account rule applies only to services "specified in regulations," so cabinet will decide which platforms are covered. Parliament would be passing the framework while the compliance perimeter is set later, outside debate. That is the "trust us" structure Geist criticises.
What committee should fix
The committee has real work to do. Four changes would keep the split coherent without gutting the bill's child-safety purpose.
- Anchor the hatred category to C-9. Tie the platform category to the criminal definition, or state expressly that a service is not required to remove lawful opinion. Two parallel definitions will drift apart.
- Write the covered-service list into the bill, or require parliamentary review of the regulations. Cabinet should not have unreviewed power to widen the perimeter.
- Separate the child-protection duties from the adult-speech duties. The strongest case for the bill is child safety. Adult political speech is where the risk to free expression is highest.
- Require transparency on over-removal. The Commission should publish error rates and appeal outcomes, not only takedown counts.
The bottom line
Splitting hate law between a criminal statute and a platform regime is a better design than C-63's single sprawling bill. But the split only works if the regulatory side does not quietly re-create a speech standard that Parliament chose to leave in the Criminal Code. As it stands, C-34 lists hatred as a category and leaves its operational meaning to a regulator and to the compliance departments of the companies being fined. The committee should close that gap before the bill returns to the House.