India’s central bank has recommended a more flexible oversight strategy for artificial intelligence systems
A Reserve Bank of India panel exploring the responsible deployment of artificial intelligence in finance has advocated for what it calls a “tolerant supervisory stance” towards AI errors, aiming to foster innovation while managing risks.
Recognising the inherently probabilistic and non-deterministic nature of AI, the report warns that a strict-liability regulatory environment could stifle experimentation and discourage institutions from harnessing AI’s full potential. Instead, the panel suggests that regulators offer measured flexibility for initial missteps—so long as financial entities maintain sufficient safeguards.
At the heart of the framework are seven guiding principles, or “sutras”, designed to anchor ethical AI use across the sector. These include: establishing and preserving trust; ensuring human oversight; encouraging purposeful innovation; promoting fairness and inclusion; upholding accountability; designing for transparency; and building systems that are secure, resilient, and energy-efficient.
Complementing these ideals are 26 practical recommendations. The panel urges the construction of enhanced data infrastructure, the creation of controlled “sandbox” environments for AI testing, and the development of indigenous AI models—particularly to help smaller players innovate responsibly.
Moreover, the framework underscores the need for flexible regulatory norms and proportionate liability frameworks. Financial institutions are advised to establish board-approved AI policies, fortify data governance, and ensure customer protection through transparency, robust grievance mechanisms, and strong cybersecurity measures.
Finally, the panel emphasises ongoing oversight—advocating continuous monitoring, public reporting, and sector-level review to maintain confidence and accountability in AI deployment across India’s financial landscape.
