The truth about scam detection solutions
Can you read minds??
Although I can predict my kids will always say “yes, please” to pudding, it’s a stretch to say, I have telepathic abilities. I’ve worked in the technology sector for 20 years, and I can strongly say that I have never seen technology understand my thinking.?
This is the problem with scam detection: how can we confidently determine why someone is taking action to prevent them from being scammed out of their money?
Fraudsters are creative and ruthless folk. They are constantly finding new ways to defraud us and get paid.?
Historically, the approach was to steal ‘the mark’s’ online banking credentials and take over their account. Banks invested a lot of time into building defences against Account Takeover and devising tactics to prevent it, learning when abnormal activity occurred on our accounts.?
With this approach, banks learn what is ‘normal’ for a user, look for anomalies, and flag these as risks. ‘Normal’ could include the device the customer uses, the location from which the transaction is made, or behavioural characteristics.?
Scams are a different kettle of fish.?
The fraudsters rightly identified that it was hard to fool these systems. They would need to replicate a device and act in such a way that it appears to be genuine. More effective, and sadly easier, to manipulate the mark to do what you need them to do.?
Fraudsters create stories to trick the customer into handing over their money.?
We have all heard about the common type of scams:
You can see that all of these scams have one thing in common. You simply need to pay or transfer money to a new beneficiary. Unfortunately, new beneficiaries are a part of day-to-day life and we all need to pay a tradesperson or transfer money to a friend with the ability to do this promptly. This is all the more necessary in a world without cash and cheques.?
The problem for banks is how to spot when the customer wants to create and pay a new legitimate beneficiary versus when they want to create and pay a fraudulent beneficiary. Somehow, the bank needs to see what external stimuli drive the transaction because, in every scenario, the customer will make the payment by themselves.?
领英推荐
So, what can be done??
There are indicators that a transaction may possibly be nefarious. Examples include:
These are just loose indicators. Many vendors provide some or all of these features, but without any guarantees that this is, in fact, a scam. They could all lead to misclassification as a scam when, in fact, the transaction is fully legitimate. This would be known as a false positive. The challenge all banks face is that there is no direct line to the customer’s mind to understand why they are making the payment.?
If the customer genuinely believes that they are doing the right thing, you can never confidently establish its legitimacy unless you can analyse the stimuli that necessitated this payment.
In October 2024, when mandatory reimbursement rules come in for scam victims, banks need more confidence that a payment is genuine, but they do not want to impact customer experience. For new payees, many banks will employ a ‘holding period’ during which investigations can be carried out and the transaction can be analysed.?
Ultimately, more genuinely defrauded customers getting refunds is never a bad thing, but it will come at a cost to customer experience.?
Yes, the scam conundrum is a very tricky problem to solve. But don’t worry, with the right technology, processes and mindset, anything can be done.?
Contact us, if you want to learn more.?
Written by Matthew Hedges