Why are organizations jumping onto the AI-enabled Identity verification bandwagon?

Bhavana Mallesh
Chief, Product Engineering

Who has not experienced being asked to show some kind of government ID, be it to receive your courier, or check into a hotel, or open a bank account or travel somewhere?

Organisations are mandated by their regulators to verify the identity of the customer, they are engaging with. Human verification is rapidly being replaced with automation using technology such as AI/ ML with OCR.

AI-enabled identity verification is the process where a technology solution steps in for a human and ensures a real human is on the other side engaging with the application, thereafter it collects government photo ID documents and finally verifies that the human and the photo IDs match.

Banks, self-service car and two-wheeler rentals were the early adopters of self-identity verification technology. The global Covid-19 pandemic has propelled the adoption of this technology and made it a norm now and that too across industry verticals, geographies and customer audience.

How does this help?

It is Convenient, Fast & Secure for the customer and impedes fraud and makes identity verification scalable for the organisation. There are several ways how a customer can invoke this technology.

It could be by clicking on a URL link that is on the organisation website, or scanning a QR code displayed on print media, via the organisation’s mobile application and embedded links in the SMS to name a few.

The primary reasons why AI technology-based identify verification became possible:

1. The genesis of OCR, Computer Vision & AI Deep learning

OCR technologies were developed as far back as the 1920s and further improved in the 1950s, to digitize printed text documents. Contemporary OCR emerged with the advent of Tessaract, an OCR engine that came inbuilt with the ability to recognise more than 100 languages, and with the ability to be trained further.

Man’s ingenuity resulted in the emergence of Computer Vision as a technology discipline under Artificial Intelligence in the 1960s. Computer Vision is the technique where a computer system is trained to mimic human vision recognition, extraction and analysis of images.

Between 2009 and 2012 researchers developed Recurrent and Convolutional neural networks for pattern recognition and machine learning. These deep learning methods help achieve recognition across different types of printed documents and high-resolution images.

A combination of Artificial Intelligence-based Computer Vision, OCR and Deep Learning augmented with carrier technologies is making document identification, facial matching, and extraction fast and secure resulting in their use for a multitude of practical solutions including identity verification. This will help organisations to significantly cut costs and improve turnaround times in customer servicing.

According to Gartner,

“The global computer vision market size was valued at USD 10.9 billion in 2019 and is expected to grow at a compound annual growth rate (CAGR) of 7.8% from 2020 to 2027.” Vision Intelligence will play a key role in the post-COVID industrial world.


2. Changing consumer behaviour

 The Internet revolution and the advent of e-commerce changed customer behaviour in using mobile devices and getting a job done with a few clicks. The recent turn of events globally has upended consumers even more and forced them to change their behaviour in how they go about almost everything they do. Going to a store or a bank or an ATM has become rife with fear and consumer beliefs have changed. They are happy if they are able to get things done from the comfort of their homes.

While there will be a day when we are looking back at the pandemic, some of these change is here to stay. Organisations realize this and are making necessary investments in technologies such as AI/ML for automating their software solutions.

Self Service applications seem to be the preferred choice for a majority of Gen Y and  Gen Z audience. It gives them the flexibility to avail themselves services 24 * 7 as and when and where they desire it from.

The above changing behaviour calls for identity verification as the first step in all remote interactions, be it for personal reasons or work. The use of these technologies to make the user experience quick, secure and glitch-free is the need.


3. Fraud Prevention

Global identity fraud is estimated to be about $46 billion and 20% of this can be attributed to new account openings. The other areas where identity fraud happen are family and elderly fraud where family members misuse identity for personal gains.

Organisations are set to combat this now with the use of identity verification technologies that provide multiple layers of protection. They are using a combination of AI/ ML based recognition of documents and automatic matching of photo IDs across demonstrated government documents.

Further this is being augmented with selfie liveness verification, two factor authentications using a registered mobile number with the government and/or email, video-based verification by the bank and Knowledge-based random questions risk assessments.

Ascertaining the identity of individuals digitally is also required during the life cycle of the use of the account during step-up transactions, updating of account details, Re-KYC to name a few. Other attributes such as location IP, usage pattern, device IPs are all making digital mode of identity verification more robust and fool proof.

4. Emerging new norms

Technology has emerged as the saviour in the COVID world. Non-contact interaction is desired both by end consumers and organisations for safeguarding the health and safety of their employees. Banks have become the proponents of identity verification, facial matching, and liveness detection technologies. Across the globe, financial institutions have embraced these technologies for survival and growth. Functions such as customer onboarding, re-KYC have become automated with Artificial Intelligence technologies.

Central Banks have also seen the need and passed regulations supporting the use of these technologies. A case in point is the Reserve Bank of India Master Directive amending KYC norms in Jan 2020 and further modified it on May 10th, 2021, to be more expansive and secure. You can read more here We have seen a similar change in regulations across Europe, US, GCC and South-East Asian countries.

There are going to be more advancements happening in the coming days. Technology providers are geared towards making their solutions to be more low code and framework based to be able to allow more customer service products and software features to be added rapidly.

There are also policies being instituted under the regulatory directives of various countries, calling for customer journey to be straight through with no breaks to discourage fraudulent activities. 

Geo-restriction checks are another feature being implemented to contain fraud. Additionally, maker/checker and concurrent audit practices are steps being implemented to ensure genuine and authentic transactions are received.

Stakeholders from internal controls and compliance also depend on ML confidence scores before authorising and approving the requisite documents.

In the coming days, as more confidence is garnered within customers and technology providers, they will be able to ensure resilient and scalable services. Use of AI/ML technologies will see explosive growth across wider use cases.


Subscribe to GIEOM blog

Social media & sharing icons powered by UltimatelySocial