Attempted banking scams increased 35% over the past 12 months, according to new data from BioCatch, with employment scams recording the sharpest rise and mobile devices now involved in nine out of every 10 reported scam sessions.

The findings come from BioCatch’s 2026 Global Scams report, drawing on data from more than 370 financial institutions across 21 countries. Those institutions collectively protect more than 760 million users and analyse around 19 billion banking sessions each month.

While a 35% increase is hardly encouraging, the rate of growth has slowed considerably from the 65% increase BioCatch reported a year earlier. The company argues this could indicate financial institutions using behavioural intelligence are becoming more effective at identifying scams before money leaves an account.

“Social engineering scams have not suddenly become less prevalent or sophisticated,” BioCatch Director of Global Fraud Intelligence Thomas Peacock said.

“If anything, artificial intelligence has lowered the barrier to entry for aspiring scammers, allowing more bad actors to create more convincing scams at a scale we’ve never seen before.“

According to BioCatch, behavioural intelligence gives banks another way to identify when a legitimate account holder may be acting under instruction or coercion, rather than simply looking for stolen credentials or obviously fraudulent transactions.

Employment scams record biggest increase

Employment scams were the fastest-growing scam category in BioCatch’s data, with the number of reported victims increasing by 258% over the past 12 months.

Purchase scams remain the most common, accounting for approximately 33% of reported scam attempts, while investment scams continue to cause the largest individual losses.

The average attempted investment scam was valued at US$6,600, almost five times the average value across all scam types.

Romance scams, meanwhile, increased by 23%. That is substantially below the 63% increase recorded in BioCatch’s previous report.

Perhaps unsurprisingly given how much banking now happens through apps, mobile devices are also overwhelmingly where scams play out.

BioCatch says 90% of reported scam attempts now originate from mobile devices, up five percentage points from last year. For comparison, mobile devices account for around 75% of traditional unauthorised fraud cases.

Banks are looking at how you behave, not just what you do

One of the more interesting sections of the report looks at how behavioural and device data can identify a possible scam before a payment is authorised.

BioCatch maps a typical banking session across five stages: login, viewing an account, adding a beneficiary, making a payment and reviewing that payment. Traditional fraud controls often focus heavily on the transaction itself, while behavioural analysis can start assessing risk much earlier.

An active phone call during a banking session, for example, does not prove somebody is being scammed. Combined with recently installed remote-access software, unusual pauses and other signals, however, it may suggest someone else is influencing the customer.

The report says around 30% of activity reported as a social-engineering voice scam involves an active phone call during the banking session.

Even typing behaviour can provide clues.

When a customer adds a new beneficiary, entering an account number in small groups with pauses between each group could indicate somebody is dictating the information over the phone.

Banks can then combine that with other factors, such as whether the recipient is new or whether the payment amount differs significantly from the customer’s usual transactions.

BioCatch’s example goes as far as analysing repeated touches and short swipes during the final payment-review screen, which the company says may indicate nervousness, distraction or someone waiting for further instructions.

In the case study included in the report, the bank detected high-risk behaviour at three separate points and displayed warnings during the session. The customer eventually backed out of the transaction before authorising the payment.

Scammers don’t necessarily need sophisticated technology

The wider report also makes the point that sophisticated-looking attacks are not necessarily the ones causing the most harm.

A section written by Operation Shamrock founder Erin West examines scam operations in Nigeria and describes an ecosystem using conventional wire transfers, gift cards, payment apps and mule accounts alongside cryptocurrency.

West argues that banks should pay attention not only to the account sending money, but also the receiving side, where funds can arrive and disappear again within the same session.

The broader message from BioCatch’s 2026 report is that stopping scams increasingly means identifying when an otherwise legitimate customer is behaving differently because someone else is manipulating them.

Passwords, MFA and transaction limits still have a role, but when the account holder is willingly pressing the buttons themselves, banks need additional signals to distinguish an unusual payment from a customer being coached through a scam.

And with attempted scams still growing 35% year-on-year, there is plenty of incentive for banks to get better at spotting that difference.

 

You can read the full BioCatch 2026 Global Scams report here.