The Monetary Authority of Singapore (MAS) is developing a framework to ensure that customers are judged fairly when AI is used to assess their credit risk, according to broadsheet The Straits Times.
As part of the first phases of a national initiative called “Veritas”, it is leading a consortium of banks and technology firms to study ways to ensure that AI-based systems employed to market financial products to customers are used appropriately.
Banks are already using AI technology to analyze data to determine their customers’ credit risk and identify relevant products to sell them, notes the MAS.
It is understood that a research report documenting the various metrics is expected to be published in the second half of 2020. This will include open-source code for the metrics that for AI practitioners to review and eventually use within their own organizations.
The Veritas consortium also aims to create a verifiable way for financial institutions to incorporate principles around fairness, ethics, accountability and transparency that MAS released in 2018 to guide firms on the responsible use of AI and data analytics, according to The Straits Times.
AI solutions and software firm Element AI says it will work together with consortium members to develop specialized metrics that will help financial institutions assess the fairness measure of their AI solutions. This will be done in collaboration with United Overseas Bank (UOB), one of the lead consortium members.
Together they will develop a set of quantifiable metrics for financial institutions to assess the extent to which data and the algorithms or models used to evaluate credit scoring for unsecured lending meet the principles of Fairness, says Element AI.
“At UOB, we uphold the highest standards in safeguarding and using data appropriately for our customers and have in place a governance framework to ensure the ethical use of data within the Bank,” said Johnson Poh, head of group Enterprise AI and Data Science, Data Management Office, UOB.
“As a founding member of the Veritas initiative and through our collaboration with Element AI, we aim to develop a set of metrics which could be used as an industry benchmark that upholds the Fairness principle in the area of credit scoring for unsecured lending.”
Photo credit: iStockphoto/Bet_Noire