Why agentic AI matters

Agentic AI is maturing and opening new possibilities for financial services. Unlike rules-based systems such as robotic process automation, agentic AI can act with greater autonomy across complex, dynamic environments. That shift means banks can automate large-scale processes that were previously impractical, affecting cost structures, operational efficiency, and customer experience.

Practical applications in banking

Banks are already applying agentic AI across a wide range of functions. These systems can handle customer service requests, automate loan approvals, and synchronize bill payments with customers’ pay cycles. They can also extract key clauses from lengthy agreements and surface actionable insights from unstructured documents. In practice, agentic AI acts as both a decision assistant and an execution layer, taking actions autonomously or escalating to humans when appropriate.

What leaders are seeing

Industry leaders highlight the technology’s potential. Sameer Gupta, Americas financial services AI leader at EY, notes that the maturing of agentic AI is ‘making it technologically possible for large-scale process automation that was not possible with rules-based approaches before.’ That capability drives measurable improvements in cost, efficiency, and the customer experience.

Murli Buluswar, head of US personal banking analytics at Citi, emphasizes that adopting new technical capabilities and redesigning operating models will separate successful firms from those that fall behind. ‘Your people and your firm must recognize that how they go about their work is going to be meaningfully different,’ he says.

Adoption and measurable impacts

Adoption is already moving quickly. A 2025 survey of 250 banking executives by MIT Technology Review Insights found that 70% of leaders report some use of agentic AI, through pilots or existing deployments. Executives point to concrete benefits: 56% say agentic AI is highly capable of improving fraud detection and 51% cite security improvements. Other notable use cases include cost reduction, efficiency gains, and enhancements to customer experience, each cited by roughly 41% of respondents.

Considerations for implementation

While the potential is significant, banks must navigate governance, risk, and integration challenges. Implementations need robust oversight to manage autonomous decision-making, clear escalation paths for human review, and careful data management to ensure privacy and compliance. Re-architecting processes and reskilling staff are often necessary to realize the full benefits of agentic AI.

About the source

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff. It was researched, designed, and written entirely by human writers, editors, analysts, and illustrators. Any AI tools used were limited to secondary production processes and underwent thorough human review.