Have you ever driven home and, when you arrived, realised that you could remember almost nothing about the journey?
You know you drove there.
You stopped at traffic lights. You negotiated junctions. You checked your mirrors. You changed speed, indicated, turned and reacted to other traffic.
You were looking through the windscreen the entire time.
But how much did you actually see?
It is something many experienced drivers will recognise.
When we first learn to drive, everything requires conscious thought. Mirrors. Signal. Manoeuvre. Gear. Clutch. Speed. Position.
It’s exhausting.
But over time something remarkable happens.
We get better.
Actions that once required deliberate thought become instinctive. We develop what we commonly call muscle memory.
Eventually, we can drive a familiar route almost on autopilot.
And that’s incredibly useful.
But it may also expose something fascinating about the way our brains deal with familiarity.
When familiarity changes what we see
This isn’t simply anecdotal.
In 2013, researchers Samuel Charlton and Nicola Starkey studied what happened when people repeatedly drove the same route in a high-fidelity driving simulator.
Participants drove the route regularly over three months, completing 20 sessions.
As the road became increasingly familiar, the mental effort required to drive it fell. Participants began describing themselves as “zoning out”, going onto “autopilot” and driving without consciously thinking about it.
More importantly, what they noticed began to change.
As familiarity increased, drivers became less likely to notice some significant changes within the environment around them.
The researchers described evidence of inattentional blindness — essentially, looking at an environment without consciously noticing everything within it. (Research Commons)
Another study published that year had an equally fascinating title:
Route familiarity breeds inattention.
Researchers Michael Yanko and Thomas Spalek compared drivers travelling familiar and unfamiliar routes.
Those on familiar routes were slower to notice pedestrians approaching the road. Their reactions to both central and peripheral events were also slower.
Perhaps most interestingly, when the researchers deliberately made the drivers concentrate on the task of driving, those differences disappeared. (Europe PMC)
Familiarity had changed their attention.
They hadn’t stopped looking.
But they weren’t necessarily seeing everything.
Change blindness
There is a related and well-established phenomenon in psychology known as change blindness.
Put simply, human beings can sometimes fail to notice changes in their environment — even changes that, once pointed out, appear obvious.
Our brains aren’t cameras continuously recording every detail around us.
We concentrate on what appears important. We build expectations about what we expect to see. And familiarity helps us process enormous amounts of information without consciously analysing every single detail.
Most of the time, that’s an extraordinary advantage.
Imagine having to consciously think about every movement required to drive a car every time you got behind the wheel.
We need familiarity.
We need routine.
We need muscle memory.
But there is another side to it.
What happens when something changes within the familiar?
And that brings me to AML.
Could compliance develop muscle memory?
Think about someone carrying out customer due diligence.
Perhaps they’ve worked in compliance for ten years.
They’ve reviewed hundreds, perhaps thousands, of customers.
• Name.
• Date of birth.
• Address.
• Identity document.
• Company.
• Directors.
• Beneficial owners.
• PEP screening.
• Sanctions screening.
• Adverse media.
• Risk assessment.
• Next customer.
• Then another.
• And another.
Eventually, just like driving, we become good at it.
We know what we’re looking for.
We know where information appears on the screen.
We recognise common false positives.
We know which documents we expect to see.
We know the questions we’re supposed to ask.
We become quicker and more efficient.
Experience is valuable.
But could that experience also create a risk?
Could compliance itself develop muscle memory?
Because if familiarity can affect what a driver notices on a road they have travelled hundreds of times, it seems reasonable to ask whether familiarity could also affect what we notice in a compliance process we have performed hundreds of times.
I want to be careful here.
The driving studies don’t prove that compliance professionals suffer from change blindness.
But they do establish something important about human behaviour.
Familiarity and repetition can affect our attention.
And AML is full of familiarity and repetition.
AML is about spotting what changed
This is where I think the question becomes particularly important.
Much of AML isn’t simply about looking at information.
It is about recognising what is different.
• Has the beneficial ownership changed?
• Has the customer’s business changed?
• Is there a new director?
• Has their geographical exposure changed?
• Has their expected activity changed?
• Has their source of wealth changed?
• Have they become a PEP?
• Has new adverse media appeared?
• Has someone been added to a sanctions list?
• Has transaction behaviour changed?
• Has something happened that changes our understanding of the risk?
In other words, we ask people to perform highly repetitive processes while expecting them to identify the one thing within that repetition which is not normal.
And perhaps that’s where we need to think more carefully.
When looking becomes a process
Compliance naturally creates processes.
We need them.
CDD must be completed.
Customers must be screened.
Risk assessments must be performed.
Reviews must take place.
Alerts must be investigated.
Records must be kept.
But there is a danger in confusing evidence that a process was completed with evidence that the risk was understood.
Consider the questions we often ask.
• Was the customer screened?
Yes.
• Was the review completed?
Yes.
• Was the alert investigated?
Yes.
• Was the customer risk assessed?
Yes.
Those answers tell us somebody looked.
They don’t necessarily tell us what they saw.
And there is a difference.
The 300th customer
Imagine reviewing your first customer.
You’re probably incredibly careful.
You read everything.
You question things.
Something unfamiliar catches your attention precisely because you haven’t seen it before.
Now imagine customer number 300.
Or 3,000.
You aren’t necessarily less competent.
Quite the opposite.
You’re vastly more experienced.
But much of the process has become familiar.
You know what normal looks like.
And knowing what normal looks like is essential in AML.
Until, perhaps, normal becomes what we expect to see.
That is where I think change blindness gives us something worth considering.
If our brain becomes accustomed to a pattern, how good are we at spotting the small change buried within it?
Particularly when we’re busy.
Particularly when there are another 50 reviews waiting.
Particularly when most of the alerts we investigated yesterday were false positives.
Particularly when we’ve known the customer for 15 years.
• “I’ve dealt with him for years.”
• “She’s always done that.”
• “It’s probably another false positive.”
• “Nothing has changed.”
Maybe.
But when did we last consciously challenge that assumption?
Alert fatigue may be part of the same problem
Take screening alerts.
If a compliance professional investigates 100 alerts and 99 are irrelevant, what does the human brain begin to expect from alert number 101?
Probably another false positive.
That doesn’t mean the person is careless.
It means they’re human.
Repetition creates familiarity.
Familiarity creates expectations.
And expectations influence attention.
The danger is that the one alert that genuinely matters arrives disguised amongst hundreds that didn’t.
Once again:
We looked.
But did we see?
So should we remove routine?
Absolutely not.
Routine is essential to compliance.
Nor am I arguing that experience is dangerous.
Experienced compliance professionals possess judgement that can’t simply be replaced by a process or a piece of technology.
The question is different.
How do we prevent familiarity from turning into complacency?
And perhaps more importantly:
How do we design compliance processes that interrupt the muscle memory when something genuinely changes?
That’s where I believe technology has an important role to play.
Not simply by automating more compliance.
But by making the important things harder for humans to miss.
If I’ve reviewed a customer before, perhaps I don’t need another system showing me hundreds of pieces of information I’ve already seen.
Perhaps I need it to tell me:
These three things have changed since you last looked.
If I’m screening a customer, perhaps I don’t need hundreds of weak matches demanding my attention.
Perhaps I need technology that does a better job of identifying the handful that genuinely deserve human judgement.
Technology shouldn’t replace the compliance professional.
It should help direct their attention to where their experience and judgement actually matter.
Looking isn’t the same as seeing
There is something slightly uncomfortable about all of this.
AML frameworks are very good at demonstrating that activities happened.
We can record when somebody completed CDD.
We can demonstrate when screening took place.
We can evidence when an alert was reviewed.
We can show when a periodic review was completed.
But ultimately, financial crime prevention isn’t about proving somebody looked at a screen.
It’s about identifying risk.
And perhaps the more familiar our compliance processes become, the more important it is that we occasionally challenge ourselves.
Are we still consciously looking for what has changed?
Or have we become so good at completing the journey that we no longer remember the road?
Because the greatest danger in AML may not be failing to look at all.
It may be looking directly at the risk — and simply failing to see it.
And if technology has a role in solving that problem, perhaps its purpose shouldn’t simply be to help us look faster.
It should help us see.