Commonwealth Bank of Australia to use AI machine learning model to detect abusive payments

It was recently announced that the Commonwealth Bank of Australia (CBA) will be developing an AI machine learning model to identify abusive payments.

Indeed, the bank has had to deal with around 106,000 transactions that were deemed threatening or abusive in the past three months, with customers having to resend these payments using appropriate messaging. The current filter is not able to further detect which payments are abusive, hence, a new AI model was created.

This model then scans patterns of behaviors, patterns of language use, and transaction types in order to find more abusive behavior. The bank also added more than 800 warning letters and account suspensions. With it, the bank has been able to identify a further 229 transactions that were potentially serious and reported around 70 customers to authorities for abusive payments.

Moreover, the bank is also launching a pilot with NSW Police so as the police can intervene and enforce protections to stop serious harassment and stalking cases. They are also trying to make sure that the victim is not put in harm’s way due to the intervention as well.