NIPFP | Homepage

(Co-authored with Lekha Chakraborty)

Artificial intelligence (AI) has rapidly become a driving force in the evolution of global and Indian financial services, fundamentally altering how institutions manage risk, optimize operations, and deliver products to consumers. The OECD defines AI as “a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.” Crucially, AI systems differ in levels of autonomy and adaptiveness, with profound repercussions for financial processes and governance.

The Influx of AI in Financial Services

Financial institutions are increasingly incentivised to use AI for their business models, owing to the opportunities for cost reduction, risk management and productivity enhancements. These institutions are currently using AI for a variety of tasks, including credit scoring to expand access to credit, reshaping client interface, automated fraud detection, stress testing, credit portfolio monitoring, automated compliance with financial regulation, forecasting and guiding investment related behaviour, to mention a few. AI also has the potential to strengthen central banks’ supervisory oversight and support them in their monetary policy mandate.

Drivers of AI Adoption in Finance

Financial institutions face unprecedented market dynamism and growing regulatory expectations, incentivizing broad AI adoption. The quest for cost reduction, risk mitigation, operational efficiency, and competitive edge has led to AI’s integration across key domains.

Credit scoring and risk assessment for financial inclusion and expansion of credit access, particularly to previously underbanked sectors, and customers with no or thin credit files.
Customer interface transformation through AI-driven chatbots and personalized advisory systems, improving service delivery and user experience.
Fraud detection, anti-money laundering (AML), and real-time monitoring of suspicious activities using advanced pattern recognition and anomaly detection.
Portfolio management and market monitoring, enabling automated stress testing, predictive analytics, and efficient compliance reporting.

Responsible AI: What It Is and What It Seeks to Address?

Concept and Necessity

The Stanford Generative AI Report and frameworks from global standard setters emphasize that responsible AI refers to AI systems designed and governed to prioritize fairness, transparency, accountability, robustness, and alignment with ethical and social norms.

Responsible AI frameworks in financial services have two primary objectives:

(1) Risk Mitigation: to mitigate risks such as discrimination, automation bias, data privacy breaches, explainability deficits, financial exclusion, and systemic instability arising from opaque algorithmic decision-making; and

(2) Outcome Driven: To ensure that outcomes produced by AI—be they credit decisions, fraud flags, or investment recommendations—are justifiable, auditable, and consistent with legal as well as social expectations.

Strengthening transparency and privacy: Clear guidance on explainability and privacy-enhancing technologies to make AI systems more accountable.
Supporting readiness and governance: Flowcharts, data privacy checklists, and implementation frameworks to assess capacity and move from principles to practice.

Financial Stability: The Core Consideration

AI’s integration affects financial stability through both amplification and mitigation of systemic risks. Precise, data-driven underwriting and automated compliance can strengthen credit discipline and reduce human error. However, risks of widespread, coordinated failures increase if multiple institutions rely on similar opaque AI models, amplifying vulnerabilities and the likelihood of correlated credit misjudgments or cyber incidents.

Central banks and supervisors can leverage AI for macroprudential oversight, market surveillance, and enforcement, but must also anticipate the heightened complexity and opacity of financial intermediation.

The Financial Stability Institute and the BIS underline the criticality of robust AI governance mechanisms to ensure resilience, prevent procyclicality, and safeguard consumer trust.

The key global benchmarks, including the OECD AI Principles, Singapore’s FEAT (Fairness, Ethics, Accountability, Transparency) guidelines, and the impending EU AI Act—stress proactive risk management, documentation, and inclusion of human oversight. The Indian regulatory landscape is moving towards similar principles emphasizing board oversight, sectoral AI inventories, comprehensive audit frameworks, and adaptive regulation.

From Mumbai to Manhattan, artificial intelligence is influencing the domain of finance: private Indian banks saw AI-related mentions in their annual reports surge six-fold between FY16 and FY23, far outpacing public-sector peers (Ghosh 2024), while globally, generative AI is already automating disclosures, powering robo-advisory, sharpening credit scoring for the previously unbanked, and transforming fixed-income analytics through synthetic data and conditional GANs (Blankespoor et al. 2025; Joshi 2025; IMF 2024a).

Yet the same technology that promises deeper financial inclusion and lower costs also amplifies systemic risks—opaque “black-box” models, recursive training data that can trigger model collapse, cyber vulnerabilities, and correlated failures if many institutions herd into identical algorithms (Cornelli et al. 2024; Finance Watch 2025; Shabsigh et al. 2023).

Recent IMF analysis warns that rapid AI-driven trading could deepen liquidity in emerging-market debt but simultaneously heighten volatility and contagion (Adrian 2024), while bibliometric surveys confirm explosive growth in applications from fraud detection and behavioral forecasting to cryptocurrency risk management, often outrunning regulatory frameworks (Milani et al. 2024; Bahoo et al. 2024).

As the EU AI Act, Singapore’s FEAT principles, and India’s emerging guidelines converge on fairness, explainability, and human oversight, the message is clear: the transformative gains from AI in finance—from cost reduction and inclusion to stronger supervision—will only be sustainable if policymakers, supervisors, and boards treat trustworthy AI not as a compliance checkbox but as a continuous journey of reflection, adaptation, and international cooperation.

Conclusion: Navigating the AI Frontier in Finance

As AI continues to permeate financial services, its promise of enhanced efficiency, inclusion, and innovation must be tempered by a resolute commitment to ethical governance and systemic resilience. The rapid adoption observed in Indian private banks, as highlighted in recent RBI analyses, underscores the urgency for policymakers and institutions to prioritize responsible AI frameworks that address bias, ensure transparency, and foster trust.

Drawing from global standards and practical tools like those in the World Bank’s report, regulators can empower financial entities to harness AI’s potential while mitigating its risks.

Ultimately, the path forward lies in collaborative action: international bodies, central banks, and private sector leaders must work together to develop adaptive regulations, invest in capacity-building, and promote explainable AI models. By doing so, we can transform AI from a potential disruptor into a cornerstone of sustainable financial stability, ensuring that technological progress serves the broader goals of equity and economic prosperity for all.

References

Abhijeet Singh is former Research Fellow at NIPFP.

(This is Part 1 of the series ‘AI in Finance’. Part 2 and 3 of this blog series on AI in Finance deal with the Global AI Regulatory Frameworks and Indian case study respectively)

Artificial Intelligence in Finance: Harnessing Innovation While Safeguarding Stability and Ethics