Finding the Needle in the Haystack
An interview with Matt Mills by Sean Murphy
Give Matt Mills a common name and he treats it as a challenge. My own name, Sean Murphy, so common in Ireland that its owner could, you might think, commit almost any crime and vanish into the crowd at least when it comes to online searches. Mills does not miss a beat. Challenge accepted, he says, before walking through exactly how Ripjar would pick the right Sean Murphy out of the thousands: an old address here, a year of birth there, a habit of holidaying in the same place. Feed the system whatever fragments you have, and an artificial intelligence layer narrows the field until one face remains. It’s far from a party trick for the team at Ripjar.
Mills became President at Ripjar in September 2025, before being asked to become CEO in January this year, and spent the previous eleven years helping to build Featurespace, the fraud detection firm eventually bought by Visa. He has a habit, he says, of arriving in markets a little too early; an augmented reality company he launched in 2012 came years before anyone wanted augmented reality. It looks like he has timed things differently with Ripjar.
To understand why, it helps to know where the company came from. Ripjar is a spin-out from GCHQ, the third and least understood of Britain's intelligence agencies. Where MI5 handles domestic threats and MI6 works overseas, GCHQ is the listening service, the one that gathers and makes sense of the world's signals and chatter. A great deal of that work is media monitoring at vast scale, sifting open information to work out what is being said and which of it matters, and it is a discipline at which GCHQ is reckoned to be among the best in the world. Ripjar's founders came out of that environment and asked a simple question: if a state can monitor the globe to inform its decisions, why can a bank not do the same to inform its own? The answer became a business that hands intelligence-grade monitoring to corporates, and Mills is candid that this lineage is the firm's real moat. That foundation, he says, is why the performance of the system sits so far ahead of the market. It is an unusually credible origin story in a sector thick with claims of proprietary cleverness, and it explains why banks, corporates, and anyone else nervous about some of the companies they work with are willing to listen.
The product at the centre of all this is unified screening - combining sanctions, politically exposed persons, watchlists and adverse media in one view - with adverse media as the youngest and fastest-moving of the four disciplines. On paper everyone agrees adverse media screening matters. Ripjar's own research, published this month, found that the great majority - 93% in fact - of financial services leaders rate it as critical. Far fewer actually do it in any serious way. The gap, Mills argues, is less about indifference than method. Firms tick the sanctions box, then type a name into a search engine and hope. The survey found that 58 per cent still rely on manual searches of this kind, and that nine in ten institutions intend to increase their investment over the coming years.
The stakes are easy to quantify. The National Crime Agency estimates that at least twelve billion pounds of criminal cash is generated in the United Kingdom each year, and Mills is quick to note that his customers operate globally, so the true figure dwarfs that. Ripjar tends to sit at the front door, working out who is opening an account before any money moves, though a growing number of clients now screen the recipients of payments too. Slow or manual checks, he suggests, are where institutions quietly lose ground, catching the wrong people late or the right people never.
What is striking is how little of this Mills attributes to regulation. Asked where the United Kingdom lags, he offers an unexpected answer: it does not, much. Britain is strong, Germany stronger still, and the explanation is largely regulatory. BaFin wrote adverse media analysis into formal guidance in 2024, and where supervisors have been directive, institutions have moved. This shows in the data which has Germany relying on manual searches the least of any nation. The outlier on the opposite end is the United States, where the current administration has warned banks against debanking customers over what surfaces online. Britain knows the argument well. In 2023 Coutts, the private bank owned by NatWest, closed Nigel Farage's accounts, and an internal dossier citing his political views later came to light. The row cost the group's chief executive her job and turned debanking into a national controversy. It is the precise tension Mills is pointing at, the line between a bank's right to choose its customers and the danger of shutting people out over what is written about them. Banks, he insists, hold both a right and a responsibility to decide who they serve.
Ripjar, now profitable and growing its recurring revenue at thirty to forty per cent a year, intends to ride that turbulence rather than be thrown by it. Mills does not expect banks to hand the problem wholesale to a frontier model; they want explainability and a traceable decision, not a black box. His ambition is narrower and more durable, to be seen as the firm that does this best.
Which brings us back to that hypothetical Sean Murphy committing crimes in Ireland, no longer quite the needle in a haystack he might have hoped to be. Somewhere in a system built by former spies, the haystack has been cleared away, and Ripjar knows how to find the needle.
