Canada's emergency scam: a spike, then a collapse

2025 Canada Ongoing

The Canadian Anti-Fraud Centre files grandparent scams under "Emergency (Jail, Accident, Hospital, Help)", and its open dataset shows a category that rose fast and then fell just as fast. Reported losses went from C$2.5 million in 2021 to C$11.6 million in 2023, then dropped to C$3.3 million in 2024 and C$1.5 million in the first nine months of 2025. It is one of the clearest examples in any national dataset of a scam category being pushed back.

Year
2025
Where
Canada
Outcome
Ongoing
Reported loss
$8.6 million
Victims
1,077
Schemes
Grandparent and family emergency scams, "Hi Mum" and family impersonation texts, Deepfake voice and video scams
Last reviewed
2026-09-06

The facts, as recorded

Why this case matters

Almost every category on this site goes up. This one went up and then came down, and that is worth understanding, because it is evidence that a scam type can be pushed back rather than merely endured.

The shape is unambiguous. Losses roughly quintupled between 2021 and 2023, then fell by 72% in a single year and kept falling. Crucially, reports and victim counts fell alongside the dollars. If this were an artefact of people reporting less, you would expect losses to hold up while reports dropped; instead all three moved together.

What was different about this scheme

The grandparent scam has a feature that most do not: it requires somebody to physically show up.

Money is collected by courier, at the door, in cash. That gives police something almost no other scam category offers — a person, at a known address, at a known time. Couriers can be arrested, and sometimes followed. Several Canadian police forces ran operations during the peak that did exactly that, and the CAFC pushed sustained public warnings at the same time.

That combination — a physical chokepoint plus a public that had been warned about a specific script — is a plausible explanation for the decline. It is not a proven one; this page is describing a correlation in reported data, not an evaluated intervention.

The caution that has to come with it

Three things could also be happening, and honesty requires naming them.

The scheme may have migrated rather than stopped. The “hi mum” text variant, which needs no courier at all, grew over the same period, and a shift between categories in the same dataset would look exactly like this.

Reported losses are a floor everywhere, and the CAFC says plainly that only a small fraction of fraud in Canada reaches it. A decline in a floor is weaker evidence than a decline in a measurement.

And 2025 covers nine months, not twelve — although even annualised it stays well below the peak.

What it tells you about the scheme

The useful conclusion is about where scams are actually vulnerable. This one was interrupted not by better filtering or better technology, but at the point where the crime had to touch the physical world. That is a general lesson: the parts of a scheme that cannot be automated are the parts that can be reached.

Sources

  1. Canadian Anti-Fraud Centre Fraud Reporting System Dataset. Canadian Anti-Fraud Centre / RCMP. Accessed 2026-09-06. Supports: Every report count, victim count and dollar figure above, aggregated from the record-level open-government extract.
  2. Report fraud and cybercrime. Canadian Anti-Fraud Centre. Accessed 2026-09-06. Supports: The CAFC's reporting role and its statement that most fraud is never reported to it.

If a legal outcome in this case has changed, please tell us and the page will be corrected. See the corrections policy.