Coinbase Wins Patent for AI-Based Crypto Compliance System

Thursday, 28/11/2019 | 19:39 GMT by Aziz Abdel-Qader
  • The system relies on an array of data points to detect risk, including users’ ages, balances, transactions volume and location.
Coinbase Wins Patent for AI-Based Crypto Compliance System
FM

A patent by US crypto exchange and wallet provider Coinbase proposes a new scoring mechanism for strengthening compliance, Risk Management , and fraud prevention through flagging non-compliant accounts. This is according to a filing published by the US Patent and Trademark Office (USPTO) on November 19.

Details about the newly published patent on the USPTO website indicate how Coinbase can build an alerts system that includes different severity levels based on their self-calculated score for each specific vulnerability and risk factors.

The proposed module relies on an array of data points to detect risk. It enters factors from user accounts into the system, then assigns a compliance score for each account based on the respective factors.

These factors include users’ ages, balances, volume of transactions, geographical location, number of devices used to access the respective account, and previous compliance reviews. It also takes into account if and how the user has verified his identity, how many changes have been made to personal details, among other 20 factors.

The next step, explains the filing, is to compare such compliance figures with a reference score to highlight those who fail to tick all boxes. The system also incorporates a freeze logic, where a security measure with which it can suspend the flagged account, halting all transactions in an adverse situation.

“A corrective action system allows for determining, for each one of the accounts that is flagged as non-compliant, whether the account is bad or good, entering the determination into a feedback system and closing the account,” it further explains.

Using artificial intelligence to drive compliance

Highlighting the possible flaws in the proposed systems, Coinbase’s solution uses AI and machine learning to help increase the efficiency of compliance tools across a variety of operations.

Since compliance alert systems based on standard technologies are triggering false positives every day, Coinbase’s module filters dozens of data elements and false alarms to improve its outcomes.

The cross-border and stateless nature of Cryptocurrencies throws up different challenges to regulators, even if they decide to devise formal guidelines to govern the asset class. As such, digital assets operators, including crypto exchanges, are increasingly turning to compliance solutions to help provide background checks on their customers as they find themselves under increasing scrutiny from regulators over possible violations of securities and AML rules.

A patent by US crypto exchange and wallet provider Coinbase proposes a new scoring mechanism for strengthening compliance, Risk Management , and fraud prevention through flagging non-compliant accounts. This is according to a filing published by the US Patent and Trademark Office (USPTO) on November 19.

Details about the newly published patent on the USPTO website indicate how Coinbase can build an alerts system that includes different severity levels based on their self-calculated score for each specific vulnerability and risk factors.

The proposed module relies on an array of data points to detect risk. It enters factors from user accounts into the system, then assigns a compliance score for each account based on the respective factors.

These factors include users’ ages, balances, volume of transactions, geographical location, number of devices used to access the respective account, and previous compliance reviews. It also takes into account if and how the user has verified his identity, how many changes have been made to personal details, among other 20 factors.

The next step, explains the filing, is to compare such compliance figures with a reference score to highlight those who fail to tick all boxes. The system also incorporates a freeze logic, where a security measure with which it can suspend the flagged account, halting all transactions in an adverse situation.

“A corrective action system allows for determining, for each one of the accounts that is flagged as non-compliant, whether the account is bad or good, entering the determination into a feedback system and closing the account,” it further explains.

Using artificial intelligence to drive compliance

Highlighting the possible flaws in the proposed systems, Coinbase’s solution uses AI and machine learning to help increase the efficiency of compliance tools across a variety of operations.

Since compliance alert systems based on standard technologies are triggering false positives every day, Coinbase’s module filters dozens of data elements and false alarms to improve its outcomes.

The cross-border and stateless nature of Cryptocurrencies throws up different challenges to regulators, even if they decide to devise formal guidelines to govern the asset class. As such, digital assets operators, including crypto exchanges, are increasingly turning to compliance solutions to help provide background checks on their customers as they find themselves under increasing scrutiny from regulators over possible violations of securities and AML rules.

About the Author: Aziz Abdel-Qader
Aziz Abdel-Qader
  • 4984 Articles
  • 31 Followers

More from the Author

CryptoCurrency