Dead Wild Roses

The Federal Court of Appeal’s January 16, 2026 decision in Attorney General of Canada v. Canadian Civil Liberties Association (2026 FCA 6) is not a cultural commentary on the Freedom Convoy. It is a narrower—and more consequential—statement about threshold: what must be true before the federal executive can lawfully reach for extraordinary powers.

The Court’s conclusion is unambiguous: “every appeal before this Court should be dismissed.” It affirms that “the declaration of a public order emergency was unreasonable,” and it adopts the word that matters in a rule-of-law system: “ultra vires.” It also states that “parts of the Regulations and Economic Order infringed paragraph 2(b) and section 8 of the Charter.” That is the spine of the judgment. Everything else is the anatomy supporting it.

What happened, in the Court’s own framing

The decision “results from the Proclamation Declaring a Public Order Emergency … issued … on February 14, 2022,” declaring “reasonable grounds to believe a public order emergency existed under subsection 17(1),” followed by “Emergency Measures Regulations” and the “Emergency Economic Measures Order” on February 15, 2022. Those instruments were deployed in the midst of what “came to be known as the ‘Freedom Convoy 2022’,” as “hundreds of vehicles” converged on Ottawa, alongside “a number of border blockades,” including the “Sweetgrass-Coutts, Alberta border crossing” and the “Ambassador Bridge.”

The Court also records the litigation path. On January 29, 2024, the Federal Court found “the reasons provided for the decision to declare a public order emergency did not satisfy the requirements” of the Act, and that some temporary measures “infringed section 8 and paragraph 2(b) … and were not justified under section 1.” The Attorney General appealed those findings; civil-liberties groups cross-appealed on peaceful assembly. The Federal Court of Appeal’s answer to the main question—was the invocation lawful and reasonable?—is no.

“Last resort” is not a slogan; it is a constraint

The most useful line in this decision is not a flourish. It is an instruction.

The Federal Court (quoted and endorsed in the appellate reasons) states: “Due to its nature and to the broad powers it grants the Federal Executive, the Emergencies Act is a tool of last resort. The GIC cannot invoke the Emergencies Act because it is convenient, or because it may work better than other tools at their disposal or available to the provinces.” It then ties that principle to factual assessment: “the evidence is clear that the majority of the provinces were able to deal with the situation using other federal law, such as the Criminal Code, and their own legislation.”

That pair of sentences does the real work. It rejects the idea—common in crisis politics—that “better” justifies “exceptional.” It anchors emergency powers to necessity, not preference.

The Court of Appeal adds a helpful nuance later: while the Act is “meant to be an instrument of last resort,” it “cannot be construed as a straitjacket” that forces Cabinet to exhaust every imaginable legal tool before acting. But that is not a concession to improvisation. It is a warning that the reasoning still has to be there, and it still has to meet “exacting” statutory demands.

Evidence, not atmosphere: “compelling and credible information”

A second throughline runs alongside “last resort”: the requirement that the executive’s belief be anchored in something sturdier than fear, political pressure, or generalized disorder.

The Court states the prerequisite in crisp terms: “the GIC must have reasonable grounds to believe, based on compelling and credible information, that threats to the security of Canada as described in section 2 of the CSIS Act existed.” And then it drops the verdict: “this evidence was lacking here.”

This is where the decision becomes a rebuke rather than a mere disagreement.

The Court rejects the attempt to stretch the statutory trigger to match the government’s policy urgency. It insists, “For the time being, we must take the Act as it reads, and not as the AGC would like it to read.” And it draws a sharp boundary around the phrase doing much of the litigation’s heavy lifting: “It would stretch beyond rationality the meaning of the words ‘serious violence’ … if they were to encompass purely economic consequences or speculative disruption of essential goods and services.”

That is a judicial way of saying: you don’t get to expand the fuse because you prefer the explosion.

When it turns to violence, the Court is equally direct. It notes that “the evidence is quite simply lacking” for serious violence against persons; and that, aside from the economic disturbance, “the only incident of violence put forward by the AGC was the seizure of a cache of firearms and ammunition at Coutts, as well as vague reports” of harassment, intimidation, and assault, plus the fact that Ottawa police were overwhelmed. “In our view, this is insufficient” to meet the “compelling and credible information” requirement for reasonable grounds.

Even where the protests were “disturbing and disruptive,” the Court concludes “they fell well short of a threat to national security.” It notes that this was “borne out by CSIS’s own threat assessment,” and it flags that an alternative assessment was requested but the Act was invoked “before it was completed.”

The implication is not subtle: emergency powers require proof-grade reasoning at the time of decision, not post-hoc narrative scaffolding.

Charter impacts: expression and financial measures

On expression, the Court endorses the Federal Court’s findings as to overbreadth and unjustified limits. It records that “the Regulations were overbroad to the extent that they criminalized the entire protest,” and that “the freedom of expression of peaceful protestors who did not participate in the actions of those disrupting the peace was infringed.” The appellate court’s own conclusion at the front end of the reasons affirms that “parts of the Regulations and Economic Order infringed paragraph 2(b) and section 8 of the Charter.”

It also disposes of the peaceful assembly cross-appeal without enlarging the rights analysis: “We are satisfied that the paragraph 2(b) infringement sufficiently accounts for the related right to assemble under paragraph 2(c), and decline to address the cross-appeal.”

On the financial measures, the Court’s concern is about sweeping power paired with loose identification and weak safeguards. It calls out “the lack of rigour contemplated in identifying individuals” who may have been subject to information-sharing provisions. Then it names the operational hazard bluntly: “The suggestion that financial institutions may have been expected to rely on information they obtained from news or social media reports or the internet … is troubling in the extreme.”

That line will survive because it captures a broader institutional truth: when the state deputizes private actors into enforcement, the state inherits the private actor’s epistemic sloppiness—unless it builds guardrails. The Court is signaling that the guardrails were not proportionate to the power.

What this decision changes (and what it doesn’t)

This judgment doesn’t settle the political argument about 2022. It does something more important: it narrows the legal escape hatch for future governments.

It teaches three lessons that will matter the next time Ottawa faces a sprawling, ugly, high-pressure public disorder:

1. “Last resort” means necessity, not superiority. “Convenient” and “works better” are not lawful triggers.
2. Economic harm is not a magic synonym for statutory violence thresholds. If Parliament wrote “serious violence,” courts will not applaud a creative rewrite.
3. Extraordinary powers demand evidentiary discipline and procedural safeguards that can withstand sunlight. “Troubling in the extreme” is what happens when you skip that.

If the government seeks further appeal, the next stage will likely fight over deference, evidentiary sufficiency, and how “exacting” the threshold must be under real-time stress. But the shape of the precedent is already clear: the Act is not a general-purpose crisis broom. It is a weapon with a trigger guard, and the Court has just tightened the guard.

References

1. Attorney General of Canada v. Canadian Civil Liberties Association, 2026 FCA 6 (Federal Court of Appeal, judgment delivered January 16, 2026; PDF).
2. Justice Laws Website, Emergencies Act (R.S.C., 1985, c. 22 (4th Supp.); enactment note “[1988, c. 29, assented to 21st July, 1988]”). (Department of Justice Canada)
3. The Canadian Encyclopedia, “War Measures Act” (notes repeal in 1988 and replacement by the Emergencies Act). (The Canadian Encyclopedia)
4. Public Order Emergency Commission (POEC), Order in Council P.C. 2022-392 (April 25, 2022) and POEC site summary of establishment. (publicorderemergencycommission.ca)