Banking Security

Robots offer helping hand to fraud investigators

Robots offer helping hand to fraud investigators, with RPA. Photo by Soultkd,
Written by Joan McGowan

RPA technology has the potential to substantially increase the productivity and accuracy of AML and fraud investigations. Story by Joan McGowan.

Banks continue to be plagued by extremely high false positives. (They report anywhere between 75% and 90% false positive rates across their fraud and AML transaction monitoring systems.) False positives trigger alerts that then have to be resolved by fraud investigators. It is a longstanding and frustrating problem. Treating false positives, which in essence are low-level alerts, require daily repetitive work by highly specialised employees. For each alert, the investigator has to toggle across multiple internal and external sites copying and pasting links and taking screen shots of information as evidence. It’s a humdrum and error-prone task. And, more crucially, it takes valuable time away from the investigation of high-risk alerts.

So it’s good to see this week’s press release from @NICE_Actimize introducing robotic process automation (RPA) into the world of financial crime investigations. Actimize has integrated RPA into its case management solution. The solution deploys attended and unattended robots onto the investigators’ screens to automate routine tasks involved in resolving alerts and cases.

I must note that @Pegasystems offers an RPA solution for repetitive back-office processes for fraud resolutions. Several other vendors have RPA for financial case management on their roadmaps. However, it’s far from being a mainstay of a bank’s financial crimes technology stack.

Being proactive

RPA is a low-cost technology that has the potential to substantially increase the productivity and accuracy of AML and fraud investigations. Actimize’s ambitious goal is to potentially increase investigator productivity by up to 50%. Whatever the percentage, the end result should be that investigators have the wherewithal to be proactive in the deterrent of suspicious activities.

It’s not a stretch for banks to deploy RPA together with machine learning and natural language processing to automate not only repetitive tasks but to carry out simple, judgement-based tasks in the resolution of cases or filing of SARs/CTRs.

Actimize’s solution offers attended and unattended robots. Attended robots are accessed by a tab on the investigator’s screen and, when required by the investigator, can support their daily tasks. Unattended robots are deployed in the background to quietly and quickly complete predefined routine tasks.

Unattended robots can perform checks on the queue of alerts to ascertain which alerts can be closed without exceptions. In most banks, this can be up to 50% of the alerts in the queue. An unattended robot can automatically call or message the customer for further information, and perform the process fulfilment activities of the outcomes of investigations, such as issuing a new card or sending a letter.

Attended robots can be used on-demand for such tasks as copying and pasting, or navigating between systems and screens to help the investigator complete the evidence-gathering processes.

As banks focus on operational efficiencies across risk platforms, it seems to me that Actimize’s goal is doable.

– This article is reproduced with kind permission. Some minor changes have been made to reflect BankNXT style considerations. Read more here. Photo by Soultkd,

About the author

Joan McGowan

Joan McGowan is a senior analyst with Celent's banking practice. Her research focus is on risk management, regulatory compliance, data security, and data analytics, with a concentration on disruptive technologies for risk transparency and data visualization.

Leave a Comment