I recently mediated a property-damage dispute over a total loss to a high-end luxury vehicle. The insurer’s position? The crash was staged. That file—and the hard questions it raised about proof—prompted this exploration on how Canadian courts actually approach civil fraud.

The result: when the cards hold, courts call fraud. When they don’t, the law won’t prop the stack. 

The Standard of Proof

Let’s get the law straight.

Canadian courts wrestled for years with inconsistent phrasing. Some decisions hinted at a quasi-elevated burden for fraud (See for example See Hanes v. Wawanesa Mutual Insurance Co. ([1963] S.C.R. 154^[1] and Continental Insurance Co. v. Dalton Cartage Co. [1982] 1 S.C.R. 164^[2]). But there is only one civil standard of proof: the balance of probabilities. Even for fraud. What changes is the expected quality of the evidence—clear, convincing, and cogent—not the yardstick itself. The Supreme Court settled this in F.H. v. McDougall, 2008 SCC 53^[3].

The Cause of Action: What You Must Prove for Civil Fraud

The leading modern statement comes from Bruno Appliance and Furniture, Inc. v. Hryniak, 2014 SCC 8^[4]. To establish civil fraud (deceit), a plaintiff must prove:

(1) a false representation;

(2) knowledge of the falsehood (or recklessness);

(3) that the misrepresentation induced the plaintiff to act; and

(4) resulting loss. Bruno also underscores that “fraud without damage gives no cause of action”. Inducement and loss matter.

Proving Intent: Inferences and the “Badges of Fraud”

Direct admissions are rare. Courts routinely infer fraudulent intent from circumstantial “badges of fraud”—implausible explanations, circular money flows, synchronized stories that wilt under scrutiny, timing that only makes sense if the plan was dishonest. The Supreme Court recently endorsed that inferential method (in the insolvency context) in Aquino v. Bondfield Construction Co., 2024 SCC 31, confirming that intent can be drawn from a mosaic of suspicious circumstances when taken together.

The lesson: in civil fraud, credibility findings and objective records often do the heavy lifting.

Recent Ontario Guidance: Misrepresenting Who Drove is Fraud

The Ontario Court of Appeal’s decision in Wong v. Aviva Insurance Company of Canada, 2024 ONCA 874 is a practical illustration. There, the insured and her mother misrepresented who was driving at the time of the collision. The Court upheld the application judge’s finding of civil fraud and confirmed that “loss” can include prejudice to the insurer’s ability to investigate and defend, not just a neat dollar figure. That is directly relevant to staged-collision disputes. Misleading the insurer about core facts can both void coverage and satisfy the loss element of fraud.

How Courts Look at Staged-Collision Allegations

British Columbia’s staged-collision trilogy in Insurance Corporation of British Columbia v Singh, 2020 BCSC 1320^[5] is a clean illustration of how courts build a fraud finding from objective anomalies. At trial, the judge found three rear-end collisions in Surrey over three months, no independent witnesses, multiple salvage vehicles tied to the same autobody shop, and pre-/post-incident connections between participants. Taken together, that mosaic justified the inference the crashes were orchestrated for payout, and ICBC recovered what it had paid. On appeal^[6], the Court of Appeal confirmed the trial judge applied the proper civil-fraud test (consistent with Bruno Appliance) and did not reverse the burden of proof. It endorsed using circumstantial “badges of fraud” and credibility findings to reach a balance-of-probabilities conclusion in staged-collision suits.

The analytical template had been sketched earlier in ICBC v. Husseinian, 2008 BCSC 241^[7], where the Court set out a staged-collision fraud framework tailored to the insurance context—requiring proof the collision was staged, that the defendant knew (or was wilfully blind) before it occurred, that a representation of an “accident” was then made to the insurer, and that the insurer paid out in consequence.

In staged-collision litigation, credibility sits at the centre, but it is rarely enough on its own. What tends to move the needle is objective material that either corroborates or contradicts the story, such as event-data-recorder downloads, telematics or infotainment logs, surveillance and intersection/storefront video, repair histories and parts orders,  cell-site and messaging timelines; and financial or relationship ties among participants. The court’s analysis is cumulative. Any one anomaly might be explainable, but a pattern of tell-tale features can support the necessary inferences about knowledge, inducement, and loss.

Can Fraud Be Decided Without a Full Trial?

Sometimes.

The Supreme Court’s access-to-justice “culture shift” in Hryniak v. Mauldin, 2014 SCC 7^[8] made clear that where the record permits fair fact-finding, summary judgment is appropriate—even in fraud cases. If the paper record (augmented by motion-judge fact-finding tools) lets the court evaluate credibility and draw the necessary inferences, the matter need not necessarily await a conventional trial. That cuts both ways in my view. Plaintiffs with a strong evidentiary package can move early, and defendants can seek early dismissal where the proof is speculative.

Practical Takeaways From the Mediation Room

Coming back to that luxury-vehicle “staging” file. The right legal test is simple, the proof is not.

Allegations of fraud will succeed on a balance of probabilities, but only with clear, convincing, and cogent evidence that checks each element from Bruno, namely specific falsehood, knowledge or recklessness, real-world inducement, and an actual deprivation—remembering Wong’s point that “loss” can include litigation prejudice. Parties who keep their focus on objective records and credible timelines generally fare better than those who rely on suspicion alone.

Bottom line

Fraud findings aren’t built with theatrics. They’re tested like a house of cards. Each essential element—falsehood, knowledge, inducement, and loss—must be a stable card.

On the civil standard, you don’t need a fortress. You need the structure to lean past 50% with clear, convincing, and cogent proof. But if one bottom card is missing—or tilted by inconsistencies, gaps in the record, or weak credibility—the whole thing can collapse and the allegation fails. In staged-collision disputes, the winning side is usually the one that stacks objective records and coherent timelines, not speculation.

When the cards hold, courts will find fraud. When they don’t, they won’t stretch the law to prop it up.