AWS: Securing FinTech with Multi-Cloud Fraud Protection

Financial services are facing a dual challenge: balancing a shift toward digital-first interaction and ensuring seamless customer onboarding, all while maintaining rigorous security.

Considering this trajectory, for AWS, its cloud infrastructure has to move beyond simple storage to offer sophisticated, AI-driven tools that protect both institutions and consumers. 

And, in the high-stakes world of fintech where identity fraud is a constant threat, AWS is harnessing its massive global scale – operating in more than 50 countries – to redefine how identity is verified in real-time.

Enabling a frictionless security frontier

When putting security hurdles in place, fintechs successfully make a most hostile environment for malicious actors – but this often creates friction for legitimate customers, too.

Enter Amazon Rekognition, an AWS service that utilises computer vision to automate image and video analysis.

What is Amazon Rekognition?

Amazon Rekognition, a deep learning-powered AWS service for image and video analysis.

It has compelling fintech applications thanks to its ability to enhance security, compliance and customer experience without requiring ML expertise. 

Banks and payment firms use its facial recognition for seamless biometric authentication, verifying identities via selfies against ID photos, reducing fraud in know-your-customer (KYC) processes while speeding onboarding.

It can also detect liveness – like blinking, for example – to block and prevent spoofing attacks, which is becoming increasingly vital as deepfakes rise in financial scams.

In transaction monitoring, Rekognition scans forged checks or documents for tampering by spotting text anomalies, altered signatures or mismatched layouts, automating anti-money laundering (AML) checks. 

As well as this, retail fintechs use Amazon Rekognition to analyse video feeds for real-time ATM surveillance, identifying suspicious activities or counting cash accurately. 

One company that models just how AWS leads in this area is Sun Finance.

The Latvian fintech – which operates across Europe, Asia, Latin America and Africa – was once bogged down by a manual, labour-intensive identity verification (IDV) process that was ripe for human error.

To overcome this, it turned to AWS to automate their workflow. By implementing Amazon Rekognition, Sun Finance transformed a sluggish verification system into a near-instant experience.

The system allows the company to compare a user’s selfie against their government-issued ID with unparalleled efficiency.

“We receive all requests within one second,” says Sergei Kiriasov, Sun Finance’s Head of Risk Technology.

“For each client, we request around 15 Amazon Rekognition operations and they typically finish processing in five to 10 seconds.”

This means that the entire document verification process now caps at a maximum of 20 seconds – a level speed vital for financial inclusion.

Thanks to the implementation of Amazon Rekognition, Sun Finance has seen its automatic approval rate reach 60% in some markets, even when dealing with low-quality images from budget smartphones.

Providing multi-layered protection through the cloud

But AWS doesn’t stop at simple face matching. To combat sophisticated spoofing attempts – where bad actors use photos or masks to mimic a real person – AWS utilises Face Liveness, a feature that verifies that the user is a real, live human being in real-time. 

AWS integrates Amazon Textract to bolster the IDV process.

While Rekognition handles the visual ‘face’ data, Textract automatically extracts printed text, handwriting and data from scanned documents.

It can accurately read text written vertically or at awkward angles, ensuring that diverse ID formats from more than 50 countries are processed without a hitch.

Resilience is core to a multi-cloud protection strategy for fintechs.

So, by applying and operating AWS’ suite of identity and fraud tools, a private repository of face vectors can be built to detect duplicate or fraudulent account creation attempts across an entire ecosystem.

AWS turns complex machine learning into plug-and-play services, removing the barrier between a customer and their finances and ensuring that financial services can be more inclusive and inherently more secure.