As digital financial services become more advanced, so too do the efforts of malicious hackers and fraudsters to crack into those valuable systems — and so too do the efforts of regulators to build better structures to avoid that abuse. To help the finance sector meet those demands, an Amsterdam-based startup called Fourthline has built a set of AI-based solutions to help with identity verification, anti-money laundering compliance, and more. Today, it’s announcing €50 million ($54 million) in funding to expand that technology.
The funding is being led by Finch Capital with other investors undisclosed. Fourthline has been around since 2013 — building its tech for five years before commercially launching in 2018 — and it has been mostly growing on its own steam, with only €70 million raised to date (which includes this latest €50 million injection).
In that time, though, growth has been impressive: its customers include N26, Qonto, Trade Republic, FlatexDEGIRO, Scalable Capital, NN and Western Union as well as marketplaces like Wish, with business growing 80% annually in the last five years. In all, it says its tech helps vet “millions” of consumers every year.
The problem that Fourthline is tackling, in a way is one perfectly suited to artificial intelligence: there are many routes that bad actors can take to exploit digital financial services, whether that’s by stealing data, by impersonating people or taking other approaches to steal money or move funds illicitly. While humans are obviously an important part of the solution to combat that, the proliferation and growing sophistication of those approaches makes the challenge increasingly insurmountable. AI-based approaches, using computer vision, machine learning and importantly immense data crunching to detect when something is not as it should be, become not just helpful but a must-have.
Fourthline’s approach currently includes around 200 checks covering areas like examining ID documents, parsing biometric data, checking records for place of residence, verifying names against sanction lists and more. Some of these might be straightforward database checks, but others are very much moving targets of complexity.
“We’ve invested heavily on the authentication side,” Krik Gunning, the CEO and co-founder of Fourthline, said in an interview. That includes, he said, “looking at ID and passports in different ways.”
It claims that this approach is able to identify 60% more fraud with an accuracy rate of 99.98%.
There are a number of startups that already exist to help fintechs and others be compliant in areas like KYC (know your customer) regulation and ID verification, but Gunning pointed out that the norm is to take a very different approach: companies built out solutions that typically lean on APIs and core technology built by third parties, which is then customized by the startup. Fourthline has taken the position that it’s better to build out its own technology from the ground up, using its own proprietary data sets, as this makes it easier to control and modify, and of course brings it better service margins, too, in the long run.
“We use our own OCR [optical character recognition] model for the number zone and another for the visual zone, to check if it’s been tampered with, and yes many have that too, but we are then also able to address the flip side of that, which is much harder: to confirm if something is authentic but also pinpoint the reason why it might or might not be so.” This involves deeper investigations of ghost images and being able to understand even the lighting used in creating an image. “We’ve invested a lot to do this.” That is one key reason why it took years for the company to launch a single product.
The company takes a very streamlined approach in what it does. There are, for now, no plans to build out services beyond KYC, AML and ID verification — so while credit scoring seems like a very adjacent market, it’s not one that Fourthline will be pursuing for now. And similarly, this is why it’s only concentrating on the financial sector.
“There are a lot of companies in this space that do a lot with AI but in all honesty it takes a lot of time and investment and knowledge and training before you can get AI models to the level you want,” he said. “The only way you can do it is by focusing. We couldn’t do this for every sector in every region in the world so this is why we do this where we can play a role which is why we focus on financial institutions in Europe.”
In a market rife with highly capitalized startups that have found it a struggle to live up to their valuations and growth expectations, Fourthline’s streamlined focus, proven returns, and importantly technology built by its own engineers are examples of what is resonating with investors now, and what is likely seen as a more healthy basis for really growing down the line.
Radboud Vlaar, managing partner of Finch Capital, described it with as much economy as Fourthline has taken with its own approach: “We’re big believers in tackling the compliance challenges in this industry through a focused growth strategy leveraging a platform approach using proprietary technology.”