Most people believe fraud happens to other people. The careless ones, the elderly, the ones who “don’t understand technology.” Statistically speaking, that belief is… optimistic.
Even within my close-knit circle, I’m often accused of scare tactics, of making a ‘mountain out of a mole hill. But when you have seen the vulnerabilities I have. Sat with victims (and criminals) and seen just how easy it was for them to do, and the victims to protect themselves against – it is hard not to talk about it as though we are in a financial crime epidemic.
So, rather than take it from me, let’s take a look at what the figures say,let’s start with the official numbers. According to the Crime Survey for England and Wales, the proportion of adults experiencing fraud has risen steadily:
That means in 2025, roughly 1 in 14 adults experienced fraud. Not 1 in 1,000. Not “rare.” One in fourteen. Picture fourteen colleagues in a meeting room. One of them was a victim this year. And that’s just what shows up in official data.
Now let’s introduce something even more uncomfortable. Victim Support has reported that 18% of fraud victims are repeat victims. Nearly one in five victims are not experiencing a one-off mistake, they are experiencing a pattern.
This is where Bayesian thinking becomes useful. Bayes’ theorem simply says: when new information appears, you update the probability. Your prior probability (2025 baseline) is about 7.1% chance of being scammed. But here is new evidence, you have already been scammed once. So your updated probability? 18%.
That means if you’ve already been a victim, your risk this year is not 7%. It’s closer to one in five. Put differently, once scammed, you are roughly 2.5 times more likely than the average adult to be scammed again.
That isn’t about intelligence. It isn’t about shame. It’s about exposure, targeting, data circulation, and behavioural signals. Fraudsters do not work randomly. They optimise.
This is just based on official statistics, let’s widen the lens. Self-report surveys often show that closer to 18% of adults say they’ve experienced fraud, not 7%. If that higher number reflects reality, then official statistics may be capturing only around 39% of actual victimisation.
Which means the majority of fraud may sit in what criminologists call the “dark figure”, crime that happens but never formally registers. Now that is related to shame and the stigma we attach to victims of fraud, amongst other reasons. But when asked anonymously in official surveys, victims will share – and the disparity between offically recorded crime and victim surveys can help us to understand what is really happening.
So here’s a fuller picture:
And yet most people still say: “It wouldn’t happen to me.” This is the fascinating part. We are comfortable discussing risk in abstract percentages. We are uncomfortable internalising it as personal probability. Seven percent feels small. One in fourteen feels different. Eighteen percent feels abstract. One in five feels close.
Fraud thrives in that perception gap, not because of ignorance, but in optimism bias. In the quiet belief that statistics apply to society, not to self. But probability doesn’t care about identity. It cares about exposure. And once you’ve engaged once, replied, clicked, transferred, or trusted, you are no longer statistically neutral. You are visible.
That’s not a moral statement; it’s a systems statement. If anything, the maths should reduce shame. Because fraud isn’t rare, it isn’t niche, and it certainly t isn’t confined to “vulnerable” stereotypes. It’s becoming structurally normalised.
The real question isn’t “Could it happen to me?” The numbers already answered that. The more interesting question is: If the risk is rising year on year, why does our perception of it remain so low? And what would prevention look like if we designed it for a world where 1 in 5 adults may experience fraud, rather than pretending it’s 1 in 500?
Let’s stay calm and do the maths. Looking back over the past three years of official statistics, fraud prevalence in the UK rose from: 5.7% in 2023, to 6.2% in 2024, to 7.1% in 2025. That offers us a trajectory to work with. If we project forward, not wildly, just conservatively (using official statistics, not self-report surveys) two scenarios emerge.
Scenario 1: Linear Growth
If fraud continues increasing by roughly 0.7 percentage points per year (as it has over the past two years):
At that point, fraud stops being an exception and becomes routine exposure.
Scenario 2: Compound Growth
If instead we assume the growth rate compounds, and recent increases suggest it might, the picture shifts more dramatically. Applying an average annual growth rate of around 11%:
Its worth just taking a moment to pause on that, one in five adults – that’s the same level self-report surveys are already suggesting today. Which raises a provocative possibility: the “future” may simply be the official statistics catching up with lived experience. But is that’s the case, will the dark figure disappear? The answer to that is very unlikely.
If 18% of victims continue to be repeat victims, at 20% annual prevalence: 20% × 18% = 3.6% of the adult population. That would mean nearly 1 in 28 adults being scammed more than once in a single year – and that figure assumes repeat victimisation stays stable.
In a world of AI-assisted targeting, data brokerage, automated phishing kits, and behavioural profiling, that 18% may not remain fixed. The more fraud scales, the more targeting precision improves. The more targeting precision improves, the more repeat vulnerability increases.
If fraud reaches 1 in 10 adults per year? 1 in 7 adults per year? 1 in 5 adults per year? We are no longer describing fringe harm, we are describing an ambient risk layer in everyday life.
Like inflation. Like data breaches. Like spam.
And here’s the real problem that we are already seeing: public perception shifts slowly, whilst fraud capability scales quickly. If prevalence doubles over a decade, will belief systems double with it? Or will most people still say: “I knew about scams. I just didn’t think it would happen like that.”
The maths suggests something uncomfortable for us all to digest; fraud is not plateauing, it is normalising. And once something normalises, prevention has to evolve from “awareness campaigns” to structural resilience. Not telling people to be careful, but designing systems that assume exposure.
Because if even the conservative model holds, the question won’t be: “Could it happen to me?” It will be: “How often will it happen, and how well are we built to absorb it?”. That’s not alarmism, it’s trajectory. And trajectories, unlike anecdotes, don’t care what we feel about them.
April in Financial Crime: The Month Enforcement Started Dismantling the Architecture May 7, 2026 For most of the last decade, financial crime enforcement has functioned as a treadmill. Investigators arrest
The Conversation About AI and Child Safety Is Half-Finished April 29, 2026 Last week, Australia’s eSafety Commissioner issued legally enforceable transparency notices to four of the largest gaming platforms in
Trust Isn’t a Feeling Any More, It’s a KPI April 17, 2026 For years, trust was treated as something soft and intuitive in business – important, but hard to define