M Rajeshwar Rao, Deputy Governor, Reserve Bank of India | Photocredit: Abhijith
The excitement of financial institutions around the benefits of AI should not be overshadowed, according to M Rajeshwar Rao, deputy governor, RBI.
In this respect, he emphasized on aspects such as governance, human-in-the-loop, maintaining data quality and safety and regulatory guardrails for new technologies, among other things.
In his keynote address with a business TV channel Banking Transformation SummitRAO noted that, under emerging technologies, artificial intelligence (AI) stands out because of the great potential for strengthening internal activities and risk management to deliver faster, more seamless customer experiences.
Although Banks have used AI in some areas of loans, there are potential usage cases for its use over other areas of the credit life cycle, he said.
clear policy
The deputy governor said that financial institutions should have an extensive strategy for AI acceptance. It must be accompanied by clear policies, risk -eetlust, critic and impact assessments, as well as ethical standards that through the organization Cascades.
“Robust monitoring and reporting mechanisms must be introduced to ensure coordination between innovation goals and institutional stability. Furthermore, it is essential in a regulated industry such as banking to understand how a model arrives in his decisions, so that the statement is a critical requirement,” he said.
There is therefore a need for financial institutions to invest in explaining AI -frameworks that offer clear, auditable reasons for loan decisions. Strong governance is central to the management of AI-driven model risk.
Rao emphasized that although AI can automate and recommend, people must be responsible for decisions.
As they accept AI for business processes, financial institutions must implement the principle of man in the loop to ensure that AI is used as an aid to support and improve human decisions and not replace them.
High quality data
Rao underlined that high -quality data is the backbone of safe and effective AI in finance.
“Although the RBI is already collecting data via supervisory reports, regulatory returns and surveys, the introduction of model risk guidelines, tailored to global best practices, will soon expand this size with data on AI models used by regulated entities,” he said.
Financial institutions must therefore adopt robust data strategies, with different, reliable indicators that reflect both the scale of AI acceptance and associated vulnerabilities.
RAO noted that the RBI recognition of the increasing use of model -controlled credit assessments and decision -making in RES, the RBI had issued a concept circular on model risk management in credit, which had established expectations about governance, validation, monitoring and accountability.
“Building on this base and recognizing the increasing use of models by the RES, not only for credit functions, but also for a wide spectrum of processes in functional and operational domains, the RBI is working on expanding the scope of these guidelines and will be over -umtils that apply to the risk management of models.
Published on September 22, 2025
#Financial #institutions #extensive #strategy #acceptance #RBI #GUV #RAO