AI is increasingly being used in high-stakes applications such as credit underwriting, fraud detection, and compliance. In these areas, errors can have direct financial, reputational, and systemic consequences. Therefore, the financial sector needs a framework or regulations to promote the technology and safeguard the stakeholders. To that effect, the RBI’s FREE AI Committee report is the first step in addressing this new tech beast. This article summarises the key insights from the first two of the four chapters. I will cover the rest in the next parts.
AI as an Enabler of Inclusion and Efficiency
A central theme in the report is financial inclusion. Traditional credit systems depend on formal financial histories and, in the process, exclude large segments of the population. AI enables the use of alternative data — such as utility payments, GST filings, mobile usage, and digital transaction patterns — to assess creditworthiness.
At the same time, AI-powered interfaces such as multilingual chatbots and voice-enabled banking can improve accessibility, particularly for users with limited literacy or language barriers.
AI is also improving efficiency across financial institutions. It can automate routine processes, strengthen risk monitoring, enhance customer engagement, and support faster decision-making. As a result, AI is becoming more mainstream in operations rather than being isolated applications.
Leveraging India’s Digital Public Infrastructure
The report highlights India’s digital public infrastructure (DPI) as a key enabler of AI adoption. Platforms such as Aadhaar, UPI, and the Account Aggregator framework provide a unified backbone for identity, payments, and data sharing. It also becomes easier for smaller institutions to tap into advanced tools. By integrating AI with this infrastructure, the financial sector can become more personalised and available in real-time.
The Need for Indigenous Models
The RBI raises concerns about reliance on global AI models trained largely on Western datasets. These models may not capture India’s linguistic diversity or economic context, leading to biased or less effective outcomes. To address this, the report emphasises the development of indigenous, task-specific models. It discussed the concept of “Trinity Models,” combining language, task, and domain. It allows AI systems to be tailored to local needs.
The Emergence of Autonomous Systems
The report also identifies the rise of autonomous AI systems. These systems can break down complex tasks, interact with other systems, and execute decisions with minimal human intervention. This represents a shift from decision support to decision automation. For example, AI systems could compare financial products across institutions and execute transactions in real time. While this improves efficiency, it also raises questions around control, accountability, and governance.
A Broad and Evolving Risk Landscape
The report highlights a wide range of risks, including bias, inaccuracy, and opacity of these models, which could make it difficult to explain decisions, particularly in areas such as credit approval. Here are some other examples of different risks:
- Errors can be amplified across large volumes of transactions, and model performance can degrade over time without proper monitoring.
- Dependence on third-party vendors introduces additional vulnerabilities, including service disruptions and concentration risk.
- Cybersecurity risks are also significant. AI systems can be targeted through techniques such as data manipulation, adversarial inputs, and deepfake fraud. At the same time, AI can strengthen cybersecurity, making it both a risk and a defensive tool.
- Widespread use of similar AI models can lead to herding behaviour and amplify market cycles. The opacity of these systems makes it harder to assess how risks may spread across institutions.
- AI systems can reinforce biases, reduce transparency, and influence financial decisions in ways that may not align with consumer interests. This raises questions around fairness, accountability, and informed consent.
Apart from them, the report also highlights non-adoption as a risk. Institutions that fail to adopt AI may face competitive disadvantages, may be less effective in countering AI-driven threats, and may fall short in expanding financial inclusion.
Trust as the Central Principle
The RBI’s framework is anchored in trust. The report states that innovation and risk mitigation must evolve together. AI has the potential to enhance efficiency, inclusion, and innovation in financial services. However, its long-term success will depend on the ability of institutions to manage risks, ensure transparency, and maintain public confidence. In this context, trust is not just a guiding principle—it is the foundation for the future of AI-driven finance.
What Comes Next?
If the first two chapters set the context, the next two focus on response and execution.
Chapter 3 examines global regulatory approaches and places India within this landscape. Chapter 4 then lays out the execution and governance framework. Built around the idea of trust, it introduces seven guiding principles and 26 recommendations across areas like infrastructure, governance, risk management, and consumer protection
Taken together, these chapters shift the conversation from why AI matters to how it should be governed.
Acknowledgement
The report has been prepared by a multidisciplinary committee chaired by Dr Pushpak Bhattacharyya (IIT Bombay), with members from NITI Aayog, IIT Madras, the Ministry of Electronics and IT, Trilegal, HDFC Bank, Microsoft India, and the Reserve Bank of India. The composition reflects the intersection of academia, policy, law, and industry required to address the complexities of AI in finance.
Follow me on LinkedIn for more information and subscribe for updates on compliance, NBFCs, BFSI, etc.