Artificial intelligence is transforming the banking industry - but not every part of the bank will feel the change equally.

According to Sopnendu Mohanty, a 20-year banking veteran and cofounder of the Singapore FinTech Festival, the back office is where the disruption will hit hardest, while front-office and middle-office roles are set to evolve and even expand.

“The front is for the customer, the middle is for the bank, and the back is for processing,” Mohanty explained. “AI changes the value of each layer — enhancing some, replacing others.”

A Career View from Inside the System

At 53, Mohanty brings rare perspective. He spent nearly two decades at Citigroup, later serving as the chief fintech officer at Singapore’s central bank, and now leads the global advisory and investment firm GFTN.

Having watched digital transformation sweep through banking, he says AI marks a new phase — one that redefines work, not just tools.

“We’ve digitized processes before. This time, we’re digitizing judgment,” Mohanty said.

The Front Office: AI as a Performance Multiplier

Roles that directly generate revenue — in investment banking, sales and trading, and corporate banking — will, he predicts, grow stronger under AI.

Large language models and generative tools are already being used to prepare pitch decks, model deals, and execute faster trades. In this space, AI is an amplifier rather than a replacement.

“AI doesn’t take away the banker’s job,” Mohanty noted. “It makes the banker faster, smarter, and better with clients.”

The integration of predictive analytics and AI-driven advisory platforms could make front-office teams more efficient and profitable, leading to more hires in data-fluent roles — from product specialists to AI relationship managers.

The Middle Office: Risk Meets Reinvention

For the middle office — responsible for risk management, compliance, and financial control — AI offers a quieter but profound revolution.

Mohanty expects this layer to benefit most in productivity, as algorithms automate regulatory monitoring, reconcile transactions, and detect anomalies in real time.

“The middle office will see efficiency soar,” he said. “AI tools will make their work faster, cleaner, and far more predictive.”

Global banks like JPMorgan, HSBC, and UBS are already deploying AI models to detect financial crime and stress-test portfolios. For regulators, the challenge will be ensuring those models remain transparent and explainable — a demand that may create entirely new job categories in AI governance and model ethics.

The Back Office: The Hardest Hit

In contrast, the back office — which spans operations, HR, IT support, and settlements — faces the sharpest contraction.

AI-enabled automation is rapidly replacing manual processes once performed by thousands of employees, from trade matching to payroll management.

“Back-office work is pure process,” Mohanty said. “AI does process better. That’s where the jobs go.”

Banks are already acknowledging this shift. At its 2025 Investor Day, JPMorgan Chase executives said AI will lead to headcount reductions and slower hiring, with operational staff expected to shrink by up to 10%.

Marianne Lake, head of consumer and community banking at JPMorgan, described the focus as “replacing repetition with reasoning.”

Even Jeremy Barnum, the bank’s chief financial officer, urged teams to “resist headcount growth where possible and prioritize efficiency.”

A Future Without Paper Walls

The automation wave is breaking down the century-old structure of banks divided into front, middle, and back offices.

As generative AI systems unify data and workflows, Mohanty argues that banks will begin to operate as integrated digital networks rather than compartmentalized departments — where software handles continuity and humans handle creativity.

“AI flattens the hierarchy,” he said. “In the future, a risk analyst and a relationship banker will work from the same dashboard. The lines between offices will blur.”

The Takeaway

Artificial intelligence won’t eliminate banking — it will reengineer it.

Front offices will grow smarter, middle offices more efficient, and back offices smaller but more autonomous.

The winners, Mohanty believes, will be institutions that treat AI not as an automation tool but as a collaborator capable of redefining value itself.

“AI won’t end banking,” he said. “It will expose what parts of banking still matter.”