Senior officials from the Reserve Bank of India (RBI) on Tuesday addressed some of the most critical issues facing India’s credit and fintech ecosystems. One was ensuring responsible AI for finance and the second was whether India’s ecosystem was ready to address the credit needs of the next 100 million borrowers.
Speaking about ensuring responsible AI for finance, RBI Deputy Governor R Rabi Sankar said there were significant risks relating to bias inherent in AI-led data, which can amplify historical discrimination in cases relating to credit profiling and hiring.
“The black box model of AI models or the lack of explanation of AI decisions makes these models not transparent, making it difficult for regulators and auditors to understand how decisions are made,” he told the audience on the opening day of the Global Fintech Fest 2025. “The absence of explainability, therefore, could constrain use of AI tools.”
“It also becomes difficult to assign responsibility when an AI model makes a harmful or erroneous decision,” he said.
Sankar said AI has moved from an enabling technology to a foundation driver for individuals and businesses in taking decisions. These assist in digital credit, underwriting and conversational banking functions. “Because financial institutions are built on trust and economies built on stability, the integration of AI in financial system is of profound responsibility.”
The RBI, in August, came out with its free-AI committee, where the report spoke about the need for a framework for responsible and ethical enablement of AI. “Where existing regulations fail to adequately cover AI-specific risks, review and amendments of guidelines should be considered,” the report said.
Also read: FM announces real-time foreign currency settlement system in Gift City
Sankar said there was a need to ensure innovation while safeguarding systemic stability. “We need this balance to ensure that AI strengthens rather than undermines financial stability. Responsible AI should not be framed as a regulatory requirement but as a matter of business ethics.”
Sankar highlighted the five Ts of responsible AI: Trust (building an AI system that upholds and enhances trust in the system); transparency (reinforces clarity and explainability in AI so that decisions can be understood, audited and questioned); training; technology (innovation) and togetherness (where the regulator, industry, academia and global partners are committed to a common goal).
The GFF 2025 also discussed another key issue for the financial ecosystem—whether India is ready for the next 100 million borrowers in terms of digital credit infrastructure—with RBI’s Executive Director P Vasudevan, YS Chakravarti, managing director and CEO of Shriram Finance, and Sashank Rishyasringa, co-founder and managing director of fintech Axio, speaking on the subject.
In terms of credit access, India’s credit to GDP ratio is 60 percent, China’s is 150 percent and US’s is 190 percent, indicating a huge headroom for credit to be provided to drive productivity.
Vasudevan said: “The present credit needs of the ecosystem are not being met at all. The difficulty is not finding the next 100 million, but it is how will we provide credit to them.” But with supply being higher than demand, it is a “happy” situation to be in. “Historically, customers first associated themselves with a specific branch of a bank, then the bank itself and now the whole banking system.”
Axio’s Rishyasringa said reaching out to the next 100 million borrowers is a challenge for the lender in terms of credit risk. “The amount of data you have to underwrite will be finite. Credit to the underserved can be done by designing other products that are less risky for lender and borrower,” he said.
First Published: Oct 08, 2025, 10:50
Subscribe Now