Solving the KYC conundrum with AI and Blockchain

In the financial world, KYC is a very manual and time consuming process. Artificial Intelligence (AI) and Blockchain combined together can streamline the process of KYC in financial services.

First, let’s take a look at what these terms mean:

BLOCKCHAINS

Blockchain can best be described as a digital database, which, unlike a traditional database, is characterized by three main features: decentralization, immutability and transparency. By using cryptography, blockchain provides a decentralized database or a “digital ledger” of transactions that everyone on the network can see. This network is basically nothing but a chain of computers that will each have to approve an exchange before it can be deemed to be verified and recorded.

Blockchain
Image: https://pixabay.com/en/blockchain-block-chain-technology-3019121/

KYC

KYC Onboarding

Know Your Customer (KYC) is a mandatory process which needs to be executed by financial institutions. This essentially boils down to understanding the background of the entity to whom you are providing financial services regardless of whether they are individuals or businesses.

For individuals, the process could be simply getting a proof of identification and address and more detailed information like understanding their wealth sources, commercial interests and status.

For companies, it is basically getting to know the actual business, history, the entity structure, leadership and shareholders. Additionally, they will need to show how the business operates and makes money.

AML

Anti-Money Laundering (AML) is the art of knowing patterns of potential illegal money flow at the transaction level. Banks get into trouble when they get caught moving money around especially if the money has illegal sources or is used for wrongful purposes, such as funding terrorism.

So, what is the problem with the KYC process today?

KYC is a tedious and expensive part of onboarding a new client. Each financial institution must perform their own KYC. This essentially means that financial institutions have a steep cost of acquiring a new customer. Let’s see how the KYC process works for an individual.

Some of the basic steps in the KYC process are:

  1. Get personal information (Name, address and source of wealth).
  2. Get proof(s) of personal information (Documents to prove name and address and source of wealth).
  3. Storing the personal information (Audit trail to show regulators that this has been done).
  4. Background checks, also known as Customer Due Diligence (CDD) (Additional checks on the background).
  5. Ongoing monitoring of customers (Check for address change, company change, status, illegal/criminal activities etc.).

All new customers applying to every financial institution must go through the same process.

Where is the bulk of the time spent in the KYC process?

  • Reading and making sense of the various connections and documents.
  • Potential multiple 2-way communication attempts with the customers to request additional information.
  • Gleaning through false positive alerts and focusing on real issues.
  • Being able to reason and look at documents like a detective to identify a potential fraud.
  • And, of course, repeating this every time there is a change to the customer including job, address, status etc.

KYC processes can be very expensive and inefficient while giving the customers a very poor user experience. It can take up to 10-20 days to onboard a customer via all the mandatory checks including numerous documents needed to be submitted and verified. This is extremely tedious for new customers and does not bode well for financial institutions since this process does not generate any revenue. To top this off, the fines that are imposed for wrong execution of the KYC process can be high. You need large teams to handle transactions that fail AML checks, and usually these have a false positive rate higher than 99%, resulting in huge inefficiencies.

So, how can AI and Blockchains help in making the KYC process more efficient?

AI plus Blockchain
Image: https://pixabay.com/en/blockchain-block-chain-technology-3012026/

As a start, creating a shared repository between the various banks and financial institutions on the blockchain and bringing customers onto it will help alleviate the customer identification step.  Blockchain creates a tamper proof, highly secure repository that automatically becomes proof for regulators as they too can become a node on the blockchain and view all information.

But how about the CDD (Client Due Diligence) step, the one that requires tons of manual effort to read, reason and analyze information from various sources? Blockchain is great since it comes with immutability, higher fault tolerance and security, shared ledger and automated smart contracts but it cannot play a role in the reasoning part of CDD. This is when we can bring in Artificial Intelligence (AI) technologies to augment blockchain to assist with the KYC/CDD challenge in a holistic way.

AI brings us object recognition, reasoning, anomaly detection and deep analytical capabilities. We can use the OCR capabilities to read and understand the various documents and store them on a blockchain based shared immutable repository that is visible to regulators. AI can perform deep analytics and reasoning on the information to detect anomalies and reduce false positives. Additionally, we can create distributed apps (dApps) and smart contracts to auto alert the participants on changes to a customer’s status or position.

So, although blockchain technology on its own might not solve the KYC conundrum completely, the combination of Artificial Intelligence (AI) and Blockchains could help in making the KYC process more palatable and efficient in the long run.

5 Replies to “Solving the KYC conundrum with AI and Blockchain”

  1. When I initially commented I clicked the -Notify me when new comments are added- checkbox and now each time a comment is added I get 4 emails with the identical comment. Is there any approach you possibly can take away me from that service? Thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *