While acknowledging the benefits of Artificial Intelligence (AI) when used to deliver financial services to groups who had been previously excluded, the Reserve Bank of India (RBI) said its use comes with risks which must be avoided by banks and other financial institutions.
In its recently-published report, an RBI panel set out a framework that was developed in December, 2024, to guide the financial sector in the use of AI.
The Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE-AI) includes seven ‘sutras’ that serve as the foundational principles of AI adoption, namely: trust is the foundation; people first; innovation over restraint; fairness and equity; accountability; understandable by design; and safety, resilience and sustainability.
Guided by these principles, the central bank’s panel recommended a forward-looking approach that includes 26 actionable recommendations under six strategic pillars.
Also Read: RBI mandates banks to clear cheques the same day
“For an emerging economy like India, AI presents new ways to address developmental challenges. But, if used without guardrails, it can exacerbate the existing risks and introduce new forms of harm,” said the report.
Multi-modal, multi-lingual AI can offer tremendous benefits, but innovation should not exacerbate risks.
To mitigate these risks, the apex bank’s panel has recommended formulation of a board-approved AI policy by regulated entities (REs) and making consumers aware when they are dealing with AI.
It has also recommended the expansion of product-approval processes, consumer protection frameworks and audits to include AI-related aspects.
Additionally, augmentation of cybersecurity practices and incident-reporting frameworks, as well as establishment of robust governance frameworks across the AI lifecycle are some of the other suggestions of the panel.
While AI can unlock new forms of customer engagement, enable alternate approaches to credit assessment, risk monitoring, fraud detection, and offer new supervisory tools, “it can lead to new risks like bias and lack of explainability, as well as amplifying existing challenges to data protection, cybersecurity, among others,” the report said.
The panel has also made several recommendations to foster the safe use of AI, including establishment of shared infrastructure to democratise access to data and compute, creation of an AI Innovation Sandbox, and development of indigenous financial sector-specific AI models.
Follow us
