Manual vs. Automated Identity Check: What’s Better and Why?
There are many challenges that any KYC method encounters; either it’s manual, automated, or combined. Dmitri Laush, CEO of omnichannel identity verification solution GetID, takes a look at each to help you decide which solution is better.
The Research and Strategy institute, Javelin Strategy, discovered that identity fraud resulted in losses of up to $16.9 billion in 2019 alone. So there is no need to explain that effective identity verification is of crucial importance in preventing money laundering and the financing of illegal activities, such as terrorism.
What are the challenges of manual ID verification?
Risk of betrayal by dishonest employees
Unfortunately, cases when the compliance staff of the company who were supposed to verify ID documents become involved in the identity theft are not rare. And if the customers find out that their vulnerable personal data was used in fraudulent schemes, they might forever lose trust in the company’s reputation.
A human workforce (salaries, computers, office rentals, etc.) will cost a lot more than any online automated software. The compensation of a well-trained compliance team will not be low at all. For example, training a compliance team (so they can stay up-to-date with the newest regulations) can cost $45/hour; that’s $225,000 a year for a 1,000-person team. The average salary of the compliance and KYC specialist is $55,333 per year, and office space per employee costs roughly $14,800 per year. In contrast, using an automated solution can reduce the cost of onboarding by 70%.
Fast onboarding of the client is what every company strives to achieve. However, with manual identity checks, the process can take days and the company can lose potential clients which wanted to use the services ASAP. They will simply go to the market to look for other options. In fact, many financial services still use manual KYC, which can take anywhere from two days to two weeks. That is how they are losing the clients who turn to competitors, which can provide faster onboarding.
One of the biggest challenges for manual ID checks is human error, because unlike machines, people get easily tired and lose their focus. In addition, errors can occur if the company doesn’t have trained and qualified compliance specialists. Lack of competence can be crucial for ID verification.
Problems with scaling
If identity checks are done manually, when entering a new market and a new country, you will need to hire a compliance staff that is competent in verifying local ID documentation and is familiar with local regulations. That can make the scaling process slow and expensive. Automated identity checks enable companies to verify and onboard new customers at scale in real time.
In reality, efficient fraud prevention requires a hybrid approach – a combination of automated and manual identity verification. This allows for an opportunity to take a closer look at the customer’s profile, perform an additional ID document check, provide an extra fraud assessment, and ask necessary questions if needed.
What are the challenges of automated ID verification?
You need a professional team of developers continuously updating the product as a whole. The product should always be user-friendly and compliant with the latest regulations for a smoother verification process.
Advantages of automated identity checks:
– This makes the onboarding process a lot easier and faster, which can help avoid losing new clients. In fact, modern solutions enable users to get verified within a couple of minutes or less depending on the complexity of the onboarding procedure.
– It can be easily scaled without hiring the whole local compliance team. Entering new markets will be simpler with automated checks.
– Easy and secure storage of sensible data provided by customers.
– Documents are verified with the help of modern technologies; MRZ recognition, OCR, and machine learning are some examples.
What do KYC providers check in the documents?
In order to successfully verify an ID document, the system would check document borders, correct color, brightness, and skews. It’ll then automatically classify the document by its type, issuing country and revision. Next, the document is checked for visual authenticity. The system checks whether the security features are present on the document, if the fonts are correct, whether the document is not a copy, whether the document is in the accordance with document template rules, and if the photo is present on the document.
After that, the MRZ code is checked for accordance with the document type. Then the MRZ is decoded into values according to the rules. Decoded values are checked for validity and compared with data extracted from the visual portion of the document.
Afterwards, the system searches for human portrait photos. This step becomes useful later when the system must compare the photo with the selfie your client provides.
Additional security features (signature, photo, ghost foto, QR code and barcode) are extracted from the document. The system checks the document’s date of expiry.
Using the document data, the system checks the user’s profile and highlights any differences it finds.
The system asks clients for selfie photos along with the ID document verification. The system will then compare the selfie photo with the portrait on the document and verify if both pictures are from the same person or not. Using computer vision, it will analyze photos and calculate the estimated percentage of similarity.
If the liveness detection check is used, then the system would check that the person is “live”. The user is then asked to perform some simple tasks, such as blinking, tilting their head, or smiling.
What is MRZ?
A Machine Readable Zone is a special zone of a document which contains a person’s data in encrypted form. The majority of passports and national ID cards have this feature. Documents containing MRZ codes are known as Machine Readable documents because only a machine can read the information saved in the document. Hence, only automated solutions will be suitable to check ID documents using MRZ.
MRZ code comes handy when checking an ID document’s authenticity. It also prevents unauthorized alteration of the IDs.
What is OCR?
Optical Character Recognition is a machine-learning tool that detects and extracts text within scanned documents, images, and even photos. Intelligent and deep OCR can even classify documents after analyzing them. It converts any type of text (even handwriting) from any images and makes it readable for any AI algorithm.
This plays a significant role in modern automated identity checks because of data extraction. And compared to manual checks, it is a much faster process.
In my opinion, automated identity checks are more suitable and advocate better to modern challenges that companies face nowadays. Because of today’s Covid-19 pandemic more and more services are shifting online, so the demand for online ID verification is growing rapidly. For now, I would advise that companies choose automated solutions, unless they are forced to do manual or combined checks by the regulatory bodies. This will allow organizations to speed up the customer onboarding process and reduce the manual labor costs without compromising regulatory compliance.
About the author
Dmitri Laush is CEO of omnichannel identity verification solution GetID.
DISCLAIMER: Biometric Update’s Industry Insights are submitted content. The views expressed in this post are that of the author, and don’t necessarily reflect the views of Biometric Update.