Banks deploy AI software layers to modernize legacy core computing systems

Banks are adopting artificial intelligence agents to work on top of core processing systems that are up to 40 years old, PYMNTS reported on October 22, 2026. Replacing legacy banking cores remains costly and disruptive, forcing employees to navigate fragmented software and perform repetitive manual tasks. Companies are now implementing modern AI software layers to handle these processes instead of undertaking full system conversions. [1]

Sutherland and FlowX.AI will deploy more than 220 prebuilt artificial intelligence agents directly over existing core infrastructure, according to PYMNTS. FlowX.AI stated that one European lender reduced manual credit handoffs by 80 percent after adopting the software, while another financial institution lowered its lending operating costs by 40 percent. The company did not name either bank. [1]

Kastle Chief Executive Officer Rishi Choudhary described the artificial intelligence tools as an action layer sitting above legacy systems, PYMNTS reported. Financial institutions are expanding access for these software agents gradually as their performance proves reliable. For example, banks use specialized collections agents to automate and process customer payments. [1]

The drive to cut technology expenses comes amid broader financial pressure on the sector. As FTMQ Business reported earlier, central banks worldwide have been preparing to raise interest rates to control persistent inflation, putting pressure on banking margins and operational efficiency. [1][2]