This article first appeared in Digital Edge, The Edge Malaysia Weekly on October 28, 2024 - November 3, 2024

Know your customer (KYC) is a crucial step in establishing trust and preventing fraud when businesses interact with their customers. The advancements in technologies such as deepfakes have, however, made traditional KYC methods insufficient.

Traditional KYC processes involve document verification and facial recognition. While these methods are useful in detecting basic forms of fraud, the process is vulnerable to more sophisticated tactics deployed by criminals, says Penny Chai, vice-president of business development Asia-Pacific at Sumsub, a full-cycle verification platform that secures the whole user journey with KYC, know your business (KYB), transaction monitoring and fraud prevention solutions.

This is because fraudulent activities such as money mule scams, romance scams and account takeovers tend to happen after the onboarding process. As such, there is a need for organisations to look beyond the KYC stage.

Artificial intelligence (AI) plays a role in both committing and combating fraud, notes Chai. Deepfakes, for example, employ AI to produce realistic fake images and videos, which are then used to impersonate individuals and commit financial and identity fraud.

Fraudsters are able to do this using AI to generate faces on both documents and selfies, and high-end graphics processing units and generative AI models to produce digitally forged documents.

On the other hand, by leveraging AI and machine learning, vast amounts of data can be analysed in real time to detect anomalies in user behaviour and identify potential fraudulent activities.

Sumsub uses signature analysis, pixel analysis, deepfake detectors and document analysis to catch fraudsters at the entry point. This helps in identifying traces of Photoshop or other graphics software, and areas that have been modified.

The deepfake detector is able to detect deepfake generators known to Sumsub. Manual forgeries and documents that are not state-issued will be able to be identified using document analysis.

But this is only the beginning, as 70% of fraud happens after onboarding, says Chai. Thereafter, AI and machine learning models are deployed for, among others, detecting deepfakes and ongoing monitoring to identify suspicious and fraudulent transactions.

“We are able to uncover hidden networks where we can see you log in in Kuala Lumpur and use the app. At the same time, if someone has actually imitated you and they are logging in in Vietnam, our fraud networks will show that this person is in two places. Immediately, it is suspicious,” says Chai.

To effectively prevent fraudulent activities from taking place, a collaborative effort involving businesses, governments and individuals is key, she says. It is essential to complement technology with robust regulations, increased public awareness and information sharing.

“We will provide as much technology as we can to be able to help fight these fraudsters and be proactive to be able to pick them up before they commit a crime. At the same time, we have many dialogues with regulators around the world about what we are watching and what they are saying,” she says.

“In fact, we are also talking to Interpol to educate them about some of what we see. At the same time, we believe it is important to also educate individuals to help understand things such as, if something is too good to be true, it probably is.”

Sumsub is partnering with Malaysian cybersecurity distributor Spectrum Edge to bolster Southeast Asia’s fraud prevention landscape and extend the company’s verification solutions across Malaysia.

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.