1. News
  2. Alerts
  3. Toronto · Phone scams
AlertToronto · Phone scams3 min read

Toronto police say scam callers now use AI to pose as your bank

Toronto police say scammers use AI to sound like your bank or carrier, and more than $50 million was stolen in the city by mid-October.

By the Dashline deskSource: iPhone in Canada
Toronto police say scam callers now use AI to pose as your bank
Photo: James Sutton jamessutton_photography, CC0
Act this week
What happenedToronto police warn that scammers are using AI to pose as banks, carriers and agencies, and over $50 million was stolen by mid-October.
Who is affectedAnyone with a phone or bank account, especially customers of banks and wireless carriers, including iPhone users in Canada.
What to do
  1. 1Treat any urgent call or text as a red flag and hang up.
  2. 2Never read out a one-time code or buy gift cards on a caller's request.
  3. 3Call back using the number on your card, bill or official website.
  4. 4Report the attempt to local police and the Canadian Anti-Fraud Centre.

Toronto police are warning that scam callers pretending to be your bank or phone carrier are getting harder to spot, and artificial intelligence is helping them. Impersonation now makes up about one in six of the roughly 17,000 fraud reports police have taken this year. By mid-October, more than $50 million had been stolen in Toronto.

What changed

Det. David Coffey of the Toronto Police Service Financial Crimes Unit told Breakfast Television that impersonation is the number one type of fraud police see. He said it covers many kinds of scams, and that it is devastating to people.

The key shift is detail. Callers may already know your name, which bank you use and even the last four digits of your card. According to the report, that information is pulled together with AI from social media, phishing and past data breaches. A caller who knows those things sounds like the real thing, but it proves nothing.

How the scam works

The script changes, but the pressure stays the same. The caller claims to be from your bank, a wireless carrier, a government office or another trusted brand. They invent an urgent threat, then present themselves as the only way out.

Coffey said the goal is to put people in an emotional state and then offer the answer to that emergency. In his words, it is emotions that the scammers target, and it is very effective. The aim is to get you to withdraw cash or hand over codes before you stop to think.

Who it affects

Anyone with a phone or a bank account can be targeted. Customers of banks and wireless carriers are the most obvious marks, including iPhone users in Canada. The same tricks show up in other Canadian warnings, from spoofed police numbers to fake texts that try to pull you into a chat.

People who hang up and check through a channel they control are in a much stronger position. Real banks and carriers will not ask you to read out a one-time code from a text. They will not demand that you withdraw cash, buy gift cards or move money into a safe account on the spot.

What to do

The alert box above lists the steps. The short version is to treat urgency as a red flag, end the call, and reach the company yourself using a number from your card, bill or an official website you typed in.

What happens next

Phone tools can cut down the noise but cannot catch everything. On an iPhone, Silence Unknown Callers and carrier spam labels help. Telus offers Scam Protection with Norton. NordVPN has brought message protection and a free call blocker called HangUp to iPhone. HangUp works with a free Nord account and sorts calls into groups such as scam or telemarketing.

What is not clear from the report is how many of the Toronto cases involved AI. Police did not break that out. Coffey also asked people to report attempts to local police and the Canadian Anti-Fraud Centre so the pattern is on record.

The takeaway

If a caller creates fear and offers a quick fix, hang up and call back on a number you trust, even if they know your details.

More articles

You should really read these.

Next step

Let’s talk