AI in BFSI – Readiness, Risk and Rewards AI is transforming BFSI with enhanced efficiency, risk assessment, and financial inclusion. However, managing bias, security, and regulations is crucial. A balanced approach with automation and human oversight will shape AI's future.

Artificial Intelligence (AI) is revolutionising the Banking, Financial Services, and Insurance (BFSI) sector by enhancing efficiency, reducing risks, and improving customer experiences. At the Tech and Innovation Summit 2025 in Bengaluru, industry leaders discussed AI's impact on BFSI, focusing on its readiness, risks, and rewards.

From risk assessment in lending to AI-driven insurance claims, companies are leveraging AI to streamline operations. However, challenges such as bias, data security, and regulatory concerns remain critical.

Readiness: AI Adoption in BFSI

Financial institutions are rapidly adopting AI to optimise decision-making and reduce manual processes. Ashwin Sekar, Chief Product and Technology Officer at InCred, highlighted how AI is transforming lending.

"InCred Finance, a risk-first, technology-driven NBFC, has built an INR 12,000 crore portfolio while maintaining profitability. Operating entirely on the cloud, InCred relies on machine learning models to assess lending risks, using bureau, banking, telco, and GPS data for probabilistic predictions. With approval rates as low as 1-10%, the company prioritises careful lending decisions."

While automation enhances efficiency, human oversight remains crucial. InCred is now integrating Generative AI to refine document analysis, reducing friction in loan approvals and improving productivity.

In insurance, AI-driven chatbots are evolving into intelligent digital assistants. Dr Anand Mahalingam, VP, Data Science at Digit Insurance, noted: "Chatbots have evolved from basic customer handshakes to intelligent solutions powered by Generative AI. With advancements like Retrieval Augmented Generation (RAG), chatbots now access extensive knowledge bases, providing faster, more accurate responses."

AI's ability to personalise insurance offerings is another key development. By leveraging agentic AI, multiple AI agents work together to provide precise policy recommendations and automate claim settlements.

Risk: Challenges in AI Deployment

Despite its advantages, AI adoption in BFSI comes with risks, including bias, security concerns, and regulatory uncertainty. Yogesh Agarwal, Founder and CEO of OnSurity, likened AI's impact on insurance to the disruption brought by Jio in telecom:

"Gen AI has transformed insurance like a 'Jio moment.' Previously, claims processing relied on rigid rule engines, making adaptation difficult. Now, AI automates decision-making, reducing human intervention. OCR engines, combined with LLMs, structure hospital bills seamlessly, enabling instant claims processing."

OnSurity has introduced a "green channel" for instant approvals on claims under INR 1 lakh, significantly reducing processing times. However, regulatory hurdles remain a challenge.

"The regulatory landscape in insurance must evolve faster with AI adoption. Ensuring audit trails for decision-making, proving unbiased outcomes, and maintaining customer protection are key."

The use of AI in underwriting and product design raises concerns about transparency and accountability. Regulators face a tough choice: should AI-driven innovations be allowed across retail insurance or restricted to large enterprises?

Rewards: AI's Long-Term Impact on BFSI

Despite challenges, AI presents immense opportunities for growth. InCred is leveraging Google's Gemini AI to enhance compliance and cybersecurity: "With Gen AI, we're enhancing regulatory compliance by automating query classification using Google's Gemini, reducing human bias. For cybersecurity, we conduct phishing simulations to improve employee awareness."

AI also plays a crucial role in ensuring financial inclusion by making credit and insurance accessible to a larger population. Dr Anand emphasised: "AI adoption is essential for business growth, but strong guardrails are crucial. Machine learning models must be transparent, unbiased, and reproducible."

Companies are implementing demographic parity assessments and using tools like LIME and SHAP to ensure fairness in AI decision-making. Regular audits and strict data governance policies are essential to prevent biases and maintain trust.

Data security is another critical concern. Sekar detailed InCred's security framework:

"We follow a structured framework to protect data, ensuring balanced investments across security areas. Our approach includes six golden signals: identifying vulnerabilities, maintaining a risk register, ensuring regulatory compliance, enhancing employee security awareness, continuously reviewing controls, and logging incidents for improvement."

As AI models continuously learn from data, a strong data infrastructure is crucial for scalability. Dr Anand stressed: "Many industries rush into AI and Gen AI, but without a strong data backbone—covering collection, storage, organisation, and governance—AI efforts remain fragmented."

For AI to drive meaningful advancements, BFSI firms must invest in data warehousing and governance to ensure quality and scalability.

AI is reshaping BFSI, offering significant rewards in terms of efficiency, personalisation, and financial inclusion. However, its risks, including bias, security vulnerabilities, and regulatory uncertainties, must be managed through strong governance frameworks. As industry leaders refine AI strategies, a balanced approach—combining automation with human oversight—will define the future of AI in BFSI.