Fig Financial has introduced an AI-based fraud screening tool designed to identify senior scam victims during the loan application process. Older applicants account for about 66% of the scam cases it flags.
The Canadian lender reviews thousands of loan applications and has identified cases in which fraudsters coach victims to take out loans on their behalf. When warning signs appear, its fraud team contacts applicants before any money is disbursed.
The launch comes amid mounting concern over financial fraud affecting older Canadians. Figures from the Canadian Anti-Fraud Centre cited by Fig show Canadians lost at least $704 million to fraud in 2025, while an estimated 5% to 10% of scams are reported.
Older adults report the largest losses of any age group, and fraud is regarded as the most common crime against older Canadians. Fig's data also shows applicants aged 65 and over make up roughly two-thirds of all scam cases identified by its internal monitoring.
Internal findings
According to Fig, Baby Boomer applicants are about 75 times more likely to be scam victims than Gen Z applicants. Saskatchewan, Manitoba and Alberta recorded the highest scam rates, at 1.5 times the rate seen in other provinces.
Nearly 9 in 10 identified scam cases, or 85% to 90%, come from urban applications, according to the lender. Victims are most concentrated in the CAD $40,000 to CAD $50,000 income bracket.
The tool, called Fig Shield, is an in-house machine learning model built around applicant behaviour. It assesses whether an applicant falls into a known scam cluster and looks for patterns associated with organised fraud targeting individuals.
High-risk clusters account for as little as 8% of daily applications. Routing those files to the fraud team has reduced time spent on manual reviews and helped staff focus intervention efforts before funds leave a victim's account.
Ardalan Shojaei, General Manager of Lending at Fig Financial, described the problem as one that often goes unnoticed until losses have already occurred.
"The heartbreak of these scams is that victims rarely know they're being exploited until it's too late, particularly vulnerable older adults," said Ardalan Shojaei, General Manager of Lending at Fig Financial.
"With Fig Shield, we can spot behavioural patterns of victims who have fallen prey to scams. This allows our fraud team to proactively reach out to applicants, helping stop a bad situation in its tracks and fundamentally changing how quickly we can protect people," Shojaei said.
Early intervention
Law enforcement officials say intervention before a payment is made remains one of the few effective ways to limit losses. Once money is transferred, recovery is often unlikely.
"The reality is that most of the money lost to these scams is never recovered, and seniors are disproportionately targeted," said David Coffey, Detective, Financial Crimes Unit, Toronto Police Service.
"Organised fraud networks are sophisticated, persistent and increasingly hard to detect. When financial institutions can identify a potential victim at the point of application and intervene early, that is one of the most effective tools we have for prevention," Coffey said.
Fig operates as a digital financial services platform in Canada, offering personal loans with structured repayment terms. The business is backed by Ontario Teachers' Pension Plan and Fairstone Bank.
The company framed the launch around growing attention on elder financial abuse, one of the most common and underreported forms of harm affecting older adults. Its data suggests the issue is not limited to isolated incidents, but appears in recurring patterns by age, geography, income and application source.
Urban applications account for the vast majority of flagged scam cases, while western provinces show the highest regional concentration. Those indicators help the fraud team decide where to focus reviews and outreach.
The model is based on behaviour shown during the application process rather than a single trigger. That approach is intended to identify people being manipulated by organised scam networks, including applicants who may not realise they are acting under a fraudster's direction.
High-risk scam clusters can be identified in a small share of daily applications, allowing staff to concentrate on a limited group of files where the likelihood of harm is greatest.