Bank of America (BofA) disclosed during its investor day that it is operating around 270 artificial intelligence (AI) and machine-learning models throughout its business functions, including back-office, customer service and fraud prevention. The bank’s leadership emphasised the aim of using AI not just for incremental cost reduction but to build a competitive “moat” through operational efficiency and innovation.
A key application cited was fraud detection: BofA reported it has halved its fraud loss rate thanks to AI-driven models that spot abnormal behaviour in real time. The bank also noted a roughly 60% reduction in service call volume, attributed to AI-enabled digital tools that handle customer interactions more efficiently. These improvements show how deeply the bank is embedding AI into its core operations rather than treating it as a side project.
On the compliance and governance front, BofA highlighted the complexity of operating AI in a heavily regulated financial environment. The bank’s international business-unit president noted that “our technology platform and those local rules and [regulations] mean that we need to be investing constantly to make sure that we’re compliant with whatever is asked for in each of these jurisdictions.” The message: scaling AI across geographies means not only building models but ensuring that data use, auditability, risk controls and local regulatory frameworks are all aligned.
Finally, BofA’s executives pointed to internal adoption and productivity gains as crucial indicators of AI maturity. For example, the bank’s Chief Technology & Information Officer said that among its roughly 18,000 software developers, the use of coding agents has yielded 20 % productivity boosts in portions of the software lifecycle. This suggests the bank is focused both on external-facing AI services and internal transformation — a dual approach that many institutions view as critical for next-generation competitiveness.
