In the last five years, generative artificial intelligence has had a profound impact on the digital landscape, sparking both incredible interest and deep concern.
The convergence of exponential increases in processing power, continuous advancements in deep learning and neural networks, and the democratisation of AI tools has fueled a creative explosion in digital media.
However, this powerful technology has also been exploited by malicious actors to create deepfakes – highly convincing simulations of individuals saying or doing things they never actually did. As deepfakes become increasingly realistic, they have the potential to pose the single biggest threat to personal identity and e-commerce security, and even more hazardously in the realm of digital banking and remote financial services, a new whitepaper reveals.
The Evolution of Deepfakes
The term “deepfake” itself reflects the technology’s foundation in deep learning and its ability to fabricate authentic-looking content by seamlessly combining and altering existing visual and audio elements.
As detailed in the whitepaper “Securing Authenticity: ZOLOZ’s Solution to Deepfake Challenges in the Digital Age“, deepfakes use a type of machine learning called deep neural networks, specifically generative adversarial networks (GANs), to create or manipulate images and videos.
According to the report, the term “deepfake” originated from enthusiast communities on Reddit in late 2017, but it has its roots in other AI advancements: while “fake” refers to the counterfeit media created, the “deep” part of the term comes from deep learning coupled with computer vision, as well as massive quantities of data and computing power.
Deepfakes have since evolved into a formidable challenge for conventional identity verification methods. The growing sophistication of deepfake technology has made it increasingly difficult for traditional eKYC solutions to detect and prevent fraud.
The Need for Multi-Layered Protection
A recent incident in Hong Kong serves as a stark reminder of the growing sophistication of deepfake-based fraud. In February 2024, criminals used a combination of stolen identity cards and AI-generated faces to deceive identity verification systems.
They successfully opened 54 bank accounts and obtained 90 loans between July and September 2023. This multi-pronged attack highlights the critical vulnerabilities in digital identity verification processes and underscores the need for robust, multi-layered protection.
To combat the ever-evolving threat of deepfakes, financial institutions must adopt a comprehensive and dynamic security stack that goes beyond minimal compliance standards. The whitepaper outlines how the ideal solution should leverage artificial intelligence and machine learning to continuously adapt to new attack types while providing rapid and accurate identity verification across a diverse range of users and devices.
It should also be configurable with localisation capability, ensuring compliance with different regional and national laws and regulations.
ZOLOZ State-of-the-Art eKYC Solution
ZOLOZ, a leading provider of eKYC technology, has developed a state-of-the-art risk control and data privacy stack that incorporates three key layers of protection: RealID, Connect, and Network.
RealID: Fast and Accurate Onboarding
RealID ensures fast and accurate onboarding by verifying ID authenticity and matching the user’s live selfie with the face on the document. This layer employs an ensemble of AI classifiers trained on an extensive library of labeled data, including official ID cards from numerous country-specific templates.
By comparing template metrics and acquired ID images, RealID can effectively detect forged or tampered documents and prevent eKYC spoofing attempts using screenshots or altered images.
ZOLOZ has the capability to customise card security feature detection based on ID design and has passed the Hong Kong Monetary Authority (HKMA) and Hong Kong police department’s “sophisticated fake ID detection test” three times with perfect accuracy.
Connect: Advanced Liveness Detection
Connect employs advanced liveness detection algorithms to prevent data injection attacks and ensure that a real human is entering the facial image data. This layer utilises a combination of techniques, such as blink detection, texture analysis, and eye region focus, to determine whether an image-based attack is occurring.
ZOLOZ Connect has passed the stringent ISO30107-3 (iBeta1/2) certification, confirming its face liveness detection eKYC capabilities meet international standards.
IdNetwork: Identifying Systemic Risks
The IdNetwork layer utilises data science graphs to identify and map duplicate authentication risks, and identity fraud risks on a large scale. By analysing historical data, including face data, device information, and user identity information, IdNetwork can effectively detect suspicious and high-risk activities, such as the same user attempting to authenticate multiple accounts or the use of tampered personal information.
A Risk-Based Approach
By adopting a multi-faceted, risk-based approach, ZOLOZ empowers its eKYC clients with layered, dynamic protection that constantly evolves to counter the latest malicious activities. The system captures suspicious and high-risk activities, allowing clients to customise their fraud risk assessments using various risk-mitigation metrics, as outlined in the whitepaper.
The aggregated results from RealID, Connect, and IdNetwork are factored into three categories of transaction results: fail, pass, or pending, with the latter typically comprising 10-15% of total transactions and requiring manual review.
Combating Deepfakes in the Digital Era
As the fintech industry continues to expand and embrace innovations like cryptocurrency and smart contracts, the need for robust and trusted security solutions has become imperative. ZOLOZ, with its cutting-edge technology and strong partnership with Alibaba Cloud, is well-positioned to deliver industry-leading eKYC solutions that can help financial institutions navigate the challenges posed by deepfakes in the digital age.
Trusted by 740 million people across Asia and honored by top e-wallet providers, traditional banks, virtual banks, lending platforms, and more, ZOLOZ is committed to providing industry-leading security that is globally trusted.
To learn more about ZOLOZ’s comprehensive approach to combating deepfake threats and securing the integrity of the organisation’s eKYC infrastructure, download the full whitepaper here. Discover how the ZOLOZ multi-layered solution can equip organisations with the secured peace of mind to offer more inclusive financial services to a wider audience.
- SEO Powered Content & PR Distribution. Get Amplified Today.
- PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
- PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
- PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
- PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
- Source: https://fintechnews.sg/94869/regtech/authenticity-in-the-era-of-deepfakes-the-multi-layered-zoloz-ekyc-approach/
