Subhankar Panda took the stage as a keynote speaker at the International Conference on Sustainable Computing and Communication Technologies (ICSCCT 2026), hosted by the University of Malta’s School of Information and Communications Technology and published in partnership with Springer, and made a point that few in the audience disputed but many remained slow to act on: The next decade of banking will depend on the extent to which institutions incorporate artificial intelligence and machine learning into their core planning, not just their back-office tools.Subhankar’s talk went beyond the familiar narrative of AI as a cost-cutting layer bolted onto legacy systems. Instead, he sees AI and machine learning as tools that should be incorporated into banks’ financial planning functions themselves – decisions about how to price risk, how to forecast liquidity and how decisions that once took committees weeks to make are now made through model minutes.“Banks that view AI as an add-on will continue to optimize yesterday’s processes. Banks that view them as planning partners will start asking better questions about tomorrow’s balance sheets. “From automation to anticipationA recurring theme in the address was the shift from AI that automates known tasks to AI that predicts unknown tasks. Subhankar noted that machine learning models are now able to stress-test portfolios against scenarios that human analysts would rarely think of constructing themselves, and are able to show early signals in trading data long before they appear in quarterly reports.This shift, he believes, has changed the job of a financial planner in the same way it has changed the job of an engineer: less time is spent crunching numbers and more time deciding what the numbers mean.“The value is not in the model that generates the answers. The value is in the banker knowing which questions are worth asking.”Trust, governance and the limits of modelsSubhankar is careful not to view the adoption of AI in banking as a purely technical issue. He spent part of his keynote discussing governance—the need for explainability in credit and risk decisions, the regulatory weight carried by financial institutions, and the reputational costs of deploying models that cannot explain their own reasoning. In an industry where a single miscalibrated model can impact customer trust and compliance risks, he suggests responsible deployment is as important as capability.This emphasis ties into a broader argument that runs throughout Panda’s recent work: reliability engineering and AI adoption are not separate conversations. His article on AI-driven test automation in enterprise delivery makes a relevant point in the software world – as systems become more autonomous, the rules for validating them must evolve equally quickly, otherwise the speed promised by AI becomes a risk rather than an advantage. Applied to banking, the same logic holds true: an AI-driven planning system is only as trustworthy as the testing and governance built around it.Calling on institutions to be patientPanda concludes by warning against viewing AI transformation in banking as a single project with an end date. Instead, he describes it as a standing capability that must be continually funded, staffed, and revisited—closer to the way institutions approach risk management than they approach software deployment.“The banks that have done it within five years are now looking at it as infrastructure, rather than banks waiting to buy a finished product.”ICSCCT 2026 attracts researchers and practitioners from the fields of computing, sustainability and applied technologies for a two-day conference at the University of Malta, the proceedings of which will be published through Springer.
