An industry veteran warns that artificial intelligence will reshape banking work from the inside out. Sopnendu Mohanty, who spent nearly two decades at Citi, said routine operations and support roles face the greatest pressure as banks speed up automation. His comments arrive as lenders roll out AI tools for compliance, reporting, and risk controls, seeking lower costs and faster processing.
Mohanty shared the warning in a recent interview, pointing to the parts of the bank that handle high-volume, rules-based tasks. He argued that the shift is already underway, and that hiring plans and training budgets need to catch up.
Who Is Sounding the Alarm?
Mohanty is best known for his work in financial technology and policy. Before that, he spent close to 20 years at Citi in senior roles. That front-row view of global banking gives weight to his assessment of where AI will bite first.
“Back office roles in banks will be hit the hardest by AI,” Mohanty said in an interview with Business Insider.
His view reflects a broader sentiment in banking circles, where leaders see automation as a path to higher margins and fewer errors. It also echoes recent surveys showing that operational functions are ripe for software-driven change.
Why Back Offices Are Vulnerable
Back office teams process payments, reconcile accounts, manage client onboarding, and prepare regulatory reports. Much of that work follows strict rules. It relies on structured data and repeatable steps.
These traits make the jobs suitable for machine learning and software bots. Banks already use tools that extract data from forms, match transactions, and flag anomalies for human review. Generative AI now drafts summaries, fills out templates, and assists with document checks.
- High-volume tasks are easier to automate.
- Structured workflows reduce the need for judgment.
- AI can handle 24/7 monitoring at lower cost.
Operations chiefs see quick returns from automating reconciliations, know-your-customer checks, and routine inquiries. The risk is that roles designed around those tasks shrink faster than new roles appear.
Impact on Workers and the Business
Job content will change before job counts do. Many teams will blend AI tools with human oversight. Staff will move from manual entry to exception handling and quality checks.
Over time, banks may reduce hiring for entry-level operations jobs. They may also reassign workers into data quality, model supervision, and control-room roles. Workers who know both banking rules and AI tools could benefit.
For banks, the gains include faster cycle times and fewer errors. But the shift carries risks: model bias, data leaks, and over-reliance on untested tools. Lenders will need to invest in validation, audit trails, and clear accountability.
Regulatory and Ethical Questions
Supervisors have warned banks to keep humans in charge of key decisions. They expect clear documentation of models and continuous testing. Regulators also watch how automation affects consumer outcomes, such as delays or wrongful account actions.
Labor groups and training advocates push for reskilling plans. They argue that banks should offer pathways into higher-value roles rather than rely on attrition. Without support, the shift could widen pay gaps and reduce mobility for support staff.
Signals From the Market
Large banks have announced AI initiatives across compliance, risk, and operations. Many are testing assistants that draft internal memos, compile case files, or summarize alerts. Consulting studies suggest a large share of banking tasks could be automated, especially in operations and support functions. That aligns with Mohanty’s warning on the parts of the bank most exposed.
Vendors now sell toolkits for document processing, sanctions screening, and real-time monitoring. Early users report shorter processing times and lower error rates. The business case is strong, especially as rates and credit costs squeeze returns.
What to Watch Next
Executives say the next year will focus on three areas: safe deployment, talent shifts, and measurable savings. Banks will refine model controls and seek faster regulatory approval cycles. They will map roles to new workflows and build training tracks for operations staff.
Clients may not see big changes at first. Most of the work happens behind the scenes. But faster onboarding, quicker dispute resolution, and fewer duplicate requests could follow as systems improve.
Mohanty’s message is blunt, and timely. AI is moving quickest where work is structured and repeatable. Back office teams fit that profile. Banks that plan early for training, governance, and role redesign will manage the change with less disruption. For workers, the safest bet is to learn the tools, aim for oversight and analysis tasks, and seek credentials that prove those skills.
