PSBs set to develop road map for deploying GenAI systems

State-run banks are set to develop a detailed strategy and road map for deploying generative artificial intelligence (GenAI) systems, with pilot use-cases to be rolled out within this financial year, said people familiar with the matter.

The lenders will also establish AI-friendly policies and governance systems, focusing on capability development, they said, adding that the initiatives are key components of the reform agenda of the state-run banks for 2024-25.

"Banks will look to strengthen the data governance and analytics functions to drive specialised use cases," said a bank executive, who did not wish to be identified.

State Bank of India, the country's largest bank, is already exploring building its own large language model to leverage the data it possesses. "Each bank will develop its own strategy and look at collaborations. This will be part of the year-end assessment of banks," the executive said. Bankers expect GenAI applications to play a significant role in credit modelling, data analytics and cybersecurity.

"Banks can use these new-age models to develop digital marketing capabilities and tech stacks to source business," the executive said, adding that other use cases include engagement through social media listening activities to track share of voice, customer satisfaction and brand health.

"The new operating models can further improve operational efficiency and the quality of back-office operations," the executive added.

The McKinsey Global Institute estimates that GenAI could add $200-340 billion in value across the global banking sector annually, or 2.8-4.7% of total industry revenue, largely through increased productivity.

Another bank executive said that one of the cornerstones of the Enhanced Access and Service Excellence reform version 7.0, launched last month, is the adoption of new-age technology and other advanced capabilities. "Lenders will focus on digitally-enabled new operating models for cost optimisation and quality excellence," he said, adding that the idea is to improve data quality and governance to increase the effectiveness of analytics use cases.

Earlier, the government had asked banks to adapt to the evolving digital landscape, seeking to ensure that the integrity of domestic financial systems remains uncompromised.