This article was first published in Express Computer.
Artificial intelligence is no longer at an experimental stage, it has impactful contributions and usage across sectors especially in India’s financial sector from chatbots to now serving as an integral part of business and operational strategy. As generative AI influences interactions, predictive models guide credit decisions, and automation discreetly identifies anomalies in real time, the pressing issue is not whether India ought to regulate AI, but rather how urgently and ethically it can implement such regulations.
India is a country of paradox where a technology like AI needs some rules and regulations in place, not to limit its functionalities but to cushion against any possible hazards. With respect to that, India has made significant progress towards responsible AI. In 2023, the Ministry of Electronics and IT (MeitY) published a draft framework advocating for voluntary ethical practices. Concurrently, the Reserve Bank of India (RBI) has provided initial guidance on AI model governance and fairness as part of its updated risk management standards for 2025. However, the regulatory oversight remains fragmented distributed among various sectoral regulators and largely interpretative.
The National Payments Corporation of India (NPCI) is currently testing AI-driven anomaly detection to maintain the integrity of UPI as it expands beyond 130 billion transactions each year. At the same time, the Unique Identification Authority of India (UIDAI) has been incorporating facial authentication models into Aadhaar services. Nevertheless, none of these implementations function under a unified ethical AI legislation or framework.
When it comes to the banking sector, this sector cannot afford ambiguity; as in these days algorithms decide from credibility of the customer for loan assessment to where frauds have occurred, credit score evaluation and customer experiences, amongst others, financial services is highly dependent on machine learning and foundational AI models. In 2024, a significant non-banking financial company (NBFC) piloted an AI-driven credit underwriting tool that mistakenly categorized over 17,000 low-income applicants as high-risk. The bias was not explicit; rather, the system was designed to prioritise digital presence and the depth of historical data, which these applicants did not possess.
Ultimately, human intervention rectified the scores, but the erosion of trust had already occurred. The initiative for financial inclusion in India relies on establishing digital trust, rather than merely ensuring digital access. As a less financially literate population begins to engage with formal banking through voice, regional applications, or assisted methods, it is imperative that AI systems needs to be monitored and held accountable by law and not just solely glorified as being efficient.
World is sitting on huge AI led cyber security epidemic and the leaders have realised urgency to regulate it. In 2024, The European Union has officially implemented its AI Act, categorising financial AI as high-risk and imposing stringent transparency obligations. Meanwhile, the United States, although still relying on executive orders and sector-specific guidelines, is swiftly enhancing expectations in critical areas such as lending and employment. Singapore’s Monetary Authority is expanding Veritas, a framework aimed at ensuring fairness and explain ability in financial AI.
In contrast, India is still in a review phase. However, considering the vast scale of our public and private digital infrastructure, any delay could lead to unintended consequences or regulatory arbitrage.
Now, what should India’s ethical framework entail?
A strong ethical framework for artificial intelligence in Indian banking must emphasise safety, accountability, and trust while fostering innovation. Essential components should encompass explain ability by design, ensuring that all significant AI decisions such as credit denials or fraud notifications should be supported by justifications that are understandable to humans. This matter is particularly pressing as more than 76% of private banks in India are implementing AI in customer-facing positions as of late 2024 (FICCI–BCG). Equally important are audits for bias and fairness, considering the well-documented existence of caste, linguistic, and gender inequalities in Indian datasets.
In the absence of compulsory evaluations, the likelihood of embedding discrimination through algorithms remains substantial. Furthermore, ethical implementation necessitates human-in-the-loop governance, which includes override mechanisms, escalation pathways, and audit logs. Informed and detailed data consent especially in public sector and rural banking is vital to progress beyond simple checkbox compliance. Finally, establishing a public registry of AI models utilized in the banking, financial services, and insurance (BFSI) sector, outlining their applications, most recent updates, and audit findings, could greatly enhance transparency.
Who can lead this movement?
It is important to setup an AI and Ethics council with members such as RBI, MeitY, SEBI, NPCI, and TRAI. These bodies will help with shared guidance, robust regulatory sandboxes, stringent laws for AI led malpractices and third party audits. Most importantly, it is important to include civil and human rights representatives and advocates to maintain balance and transparency. We must also invest in educating the children, adults in the country about AI and the ethics to be used in AI many research institutes like IITs, NITs should be consulted for a systematic curriculum.
We must transition from a scale to high trust society. India has demonstrated its capability to develop digital systems at scale. However, in the realm of banking, AI must be designed with trust in mind. The regulations we establish now will influence how 1.4 billion individuals engage with algorithms that determine their economic destinies. Regulation should not be viewed as a hindrance to innovation; rather, it is a necessary foundation for sustainable and inclusive innovation.
Just as India pioneered the world with digital public infrastructure such as UPI and Aadhaar, it now has the chance to set a global standard for responsible AI governance in the financial sector. Let us create AI that is not only intelligent but also responsible.
