With the proliferation of digitization across business, government and social environments, individuals have become more dependent on their digital identity for social interactions and business engagements. However, this has also resulted in the rise of frauds occurring due to theft of digital identity and impersonation of individuals. Fortunately, it is possible to improve detection of online frauds and secure privacy of user information by leveraging decentralized ledger.
Online frauds often cause loss of money and grief for individuals who are impersonated or whose services are consumed through online frauds. Banks also incur cost in conducting investigation on reported online frauds. Also, there is an impact on customers’ confidence in the banks’ ability to detect and prevent online frauds.
With rapid advancements in technology, fraud detection has become a complex process as impersonators explore various latest technologies. This complex process is led by anomaly detection. There are five layers of anomaly detection build into fraud detection systems –
- Layer 1: Endpoint Authentication
- Layer 2: Anomaly with Session
- Layer 3: Anomaly with Accounts
- Layer 4: Anomaly with Multichannel of the same Account
- Layer 5: Anomaly with Multichannel and Multiple Accounts
The capabilities for Layer 4 and 5 can be enhanced by utilizing DID (Decentralized Identifiers) for anomaly detection process. These DID based systems blend well with the existing systems and provide additional benefits like compliance to regulations like GDPR, increased efficiency of the anomaly detection systems and enhanced customer privacy for the financial institutions.
What are DID based anomaly detection systems?
DID systems employ Zero Knowledge Proof process and DID documents saved in the ledger for anomaly detection. Customer identity and spend patterns are converted as DID documents. The DID solution is built on an ecosystem of peers / trusted bodies and customers, all of whom will play the role of issuers, verifiers and holders.
This additional layer of DID based defense will help in:
Prevention: Leveraging DID and identifying patterns unique to a digital transaction in conjunction with the existing fraud detection systems will enable capture of patterns called as Indicators of Fraudulent Transactions (IOFT) and anchor the IOFT in a DID based distributed ledger.
Detection: Leveraging DID allows us to identify if valid proofs are associated with each transaction making use of sophisticated ZKPs (Zero knowledge Proofs) for anomaly prevention.
Revoke: Online transactions will be verified by ZKPs as part of the prevention process and on detection of a fraudulent transaction, the ZKPs will deny the authorization for the transaction thereby revoking the transaction request initiated from the source.
Detection of online frauds using Decentralized Identity Management in a Bank