Back in 2015, three brothers came up with an idea for a verification engine tailored specifically to the customer’s needs. Back then, the Sever brothers wanted to meet a need they saw in the market. Since then, the need for powerful KYC/AML solutions has only increased. Six years after its founding, Sumsub has more than 700 clients, 100 team members, 6 offices across the globe, and more than 6500 types of documents that they verify for their global clientele. Sumsub offers a total of 76 solutions ranging from KYC/AML, corporate verification, and liveness technology, to chargeback protection and cryptoasset transaction monitoring. Their solutions are customized to the requirements of local regulators and they currently cover 220 countries and regions. We had a chance to talk to Jacob Sever, the Co-founder and Chief Product Officer about Sumsub’s mission and the current KYC/AML environment.
The Idea Behind Sumsub And Its Competitive Advantage
“When my older brother, Andrey, told me that only 10% of fraudsters were caught in cases of insurance fraud, we saw that we could solve this issue with our own “anti-photoshop” technology,” Sever says. “We were already experimenting with neural networks and machine learning, which helped us get started on building a system for exposing the fake documents used by fraudsters to fool insurance companies.”
“Later on, our friends at Cryptopay asked us to develop a Know Your Customer system for them, which meant building a sophisticated and automated identity verification solution. This solution needed to not only expose fake passports and IDs, but also check them across databases and match selfies with passport photos—photos—not to mention that all these processes had to be user-friendly and compliant with UK regulations,” he continues.
Due to the high demands of the identity verification solution, the Sever brothers have been joined by Vycheslav Zholudev who took on the role of Chief Technical Officer. As Sever says, because of his expertise and help, Sumsub could take a huge step forward and accelerate the development of the solution that quickly gained popularity.
Since then, the Sumsub team has been working tirelessly on expanding the company, both in terms of improving the quality of the existing products and developing a range of new services that could help the clients meet the complex regulatory demands. As Sever says himself, “our passion for protecting companies from fraudsters, regulatory fines, and bad conversion is only getting stronger.”
Sumsub is one of many KYC/AML Compliance startups globally. With hundreds of KYC Solutions companies across the world and more appearing every year, distinguishing oneself and creating a competitive advantage in the market presents a big challenge. When asked how his company is differentiating itself in the sector, Sever says:
“We don’t provide a partial solution for our clients; we cover their needs entirely. From KYC, KYB, and payment fraud prevention to crypto monitoring and video verification, we prevent the need to shop around for separate, often costly solutions. That’s what makes us different. And if there’s something a client needs incorporated into their platform to stay compliant, they can send us a feature request—and we’ll figure out how to help them in the most efficient way. This is usually how we expand the functionality of our solution.”
Liveness Detection Software That Enhances Fraud Defenses
Moreover, Sever emphasizes that their customers will always be a number one priority. They try to listen to any issues the clients may have and provide instant support through messengers. There are no waiting lines, ticket systems, or waiting days to hear back. The customers can contact not only the team but also the founders directly.
Sumsub has a range of products, from video verification and crypto compliance to payment fraud prevention. The newest feature launched by the startup is a liveness detection software. Called “Prooface”, the software was built to eliminate popular conversion setbacks and further enhance fraud defenses.
“First of all, Prooface collects biometric data from multiple angles, which makes it more effective against any modern fraud vectors, such as presentation attacks, replay attacks, API injection, and deep fakes,” Sever says. “Second, this liveness check is very easy to complete, as there are no overburdening tasks. We simply ask users to turn their head in a small circle.”
Sever points out that their newest feature was met with the huge enthusiasm of their global customer base. Conversion has increased by an average of 20%, with some of the customers seeing a conversion as high as 44%. Moreover, positive results were observed among not only the finance industry, but also digital banking, gaming, insurance, gambling, and many other industries.
The Challenge Of Complying With Continuously Changing AML Regulations
In the KYC/AML sector, effectively detecting fraud is a major aspect. However, an equally important part is complying with the continuously changing AML regulations. In order to stay compliant, businesses need to meet country-specific KYC requirements, perform AML screening, conduct ongoing monitoring, implement a series of measures against fraud, and establish reporting procedures. As Sever notes, in several countries such as Germany, Austria, or Switzerland, the process is even more complicated as businesses have to perform “real-time video verification with compliance officers present during the checks.”
“Compliance is extremely tricky, especially for those who operate internationally—and especially now, when regulators pay much more attention to what you’re doing,” Sever says.
“All of these layers make compliance especially complicated, so larger companies can’t simply rely on manual user processing or building their own in-house compliance systems. So it’s better to seek professional help to maximize the efficiency and accuracy of this process. After all, compliance fines are not something you want to mess around with,” he adds.
Failing to comply with the regulations can lead to serious consequences. Although the size of the fine depends on the violation and the local regulations, they can go as high as a couple billion. Moreover, as Sever points out, it is not only about the financial penalty but also about the reputational damage and the consequences of undetected fraud. He uses the case of Deutsche Bank as an example to illustrate the severe consequences of not complying with the regulation. Recently, Deutsche Bank was faced with a hefty $150 million fine for inadequately monitoring ties to Jeffrey Epstein. Apart from the large fine, the case resulted in terrible publicity and decreased customer trust.
Getting In Trouble For Outdated Age Verifications
Another aspect that regulators are paying significantly more attention to in recent years is age checks. A lot of businesses are still using outdated methods of age verifications, relying on a pop-up box where customers can tick off their age. Last year, Amazon got into trouble with the regulators for the inadequate age checks.
“The problem is that there are no specific guidelines; companies are simply asked to implement ‘adequate methods’ to determine their purchaser’s or user’s age, which isn’t very helpful. Current verification methods are either very naive and trust-based,” Sever says.
“Software like Sumsub ensures proper verification but shortens the time it takes to get verified, lessening the burden on your compliance officers and making the onboarding experience much more enjoyable for users. For example, our system extracts date of birth info from ID documents in seconds, checking the authenticity of the document and determining if the age is appropriate for the location/service they are trying to access. This entire check takes around a minute—much faster than the manual process, which takes over 10 minutes.”
6AMLD And Its Impact On AML Procedures
A few months ago, the 6th Anti-Money Laundering Directive (6AMLD) was released, leaving a lot of businesses in the European Union wondering how it will affect their existing AML procedures. However, Sever says that the overall changes are minor.
“First, the term ‘money laundering’ now includes acquisition, concealment, and distribution of assets that result from criminal activity,” he says. “This can include artworks or real estate. There are also two new recognized areas where companies will need to detect both money and property laundering. These are cyber and environmental crime—such as the illegal wildlife trade and air pollution.”
“To deal with this, I recommend enhancing your transaction monitoring, Customer Due Diligence, and adverse media screening procedures. Also, don’t forget to make sure that your team is aware of these changes and is ready to act on violations,” he continues.
When talking about identity verification, a lot of our readers are mentioning deepfakes. We asked Sever how easy it is for fraudsters to fool face verification.
“Creating a good deepfake takes a lot of time and money,” he says. “Very few fraudsters currently have all the necessary resources for creating proper deepfakes—which leads me to believe that concern over deepfakes is overhyped. Most deepfakes cannot pass quality face identification systems, like liveness. Although deepfake technology could bring harm in some other ways—like impersonation, misinformation, and attacks on one’s reputation—existing deepfakes are mostly suitable for entertainment.”
“I’ll also admit that not all face verification technologies are equal—and some could be fooled quite easily. A good face verification or liveness solution should be able to easily differentiate between real faces and artificial objects, such as masks.”
Sever points out that a trustworthy verification solution that can analyze image depth, eye reflections, skin texture, and blood flow, all types of fraud will be detected.
To learn more about Sumsub and its mission, take a look here
Do you want to read the other exclusives?
EXCLUSIVE: How Can Every Company Be A FinTech? – Interview With Tarun Bhasin, The CEO of Kunai
Joe Robinson, CEO Of Hummingbird, Discusses FinCEN Leaks, FinTech, And AML
