Summary of this article
Responsible AI can help close India’s insurance gap.
Transparency and fairness must guide AI adoption.
Human oversight remains crucial for complex claims.
By Parag Raja
Over the past two years, I have watched artificial intelligence (AI) move from a talking point in strategy discussions to a force running quietly through nearly every part of how businesses operate – be it assessing risk, how quickly the claim is being settled, as well as how a first-time customer discovers the right cover for their family. For decades, life insurance in India has been built on trust – a promise made today, honoured decades later, and now AI is reshaping how that promise gets made and kept.
The insurance industry stands at an inflection point, and the larger question is no longer whether AI will transform the life insurance sector, but whether it will do so responsibly. I spend a lot of time asking a simple question, are we using technology in a way our customers would trust if they could see exactly how it worked?
The stakes are high and so is the responsibility. Looking at the numbers, India’s insurance penetration stood at just 3.70 per cent in FY25, roughly half the global average of 7.30 per cent, according to the latest annual report by the Insurance Regulatory and Development Authority (Irdai). Life insurance penetration actually slipped to 2.7 per cent from 2.8 per cent the previous year, even as premium income grew 6.73 per cent to touch Rs 8.86 lakh crore in 2024-25. This paradox – rising premiums, falling coverage – tells us something important: growth alone will not close India’s protection gap. With the onset of technology and if AI is used responsibly, it can deliver all three – trust, accessibility and precision which can help close the gap.
Across financial services, AI has moved past the pilot stage and is now built into daily operations. Insurers are using AI to underwrite faster, detect fraud, personalise products, and resolve claims in hours rather than weeks. Companies are now integrating AI-powered chatbot that helps frontline teams resolve underwriting queries in real time, alongside a digital sales assistant that generates instant quotes, and cross-verifies data across sources to catch fraud before it reaches a policy. Few companies have also deployed AI-enabled voice systems to intimate maturity claims and are experimenting in this regard.
AI is also used for extraction of proposal-form data to speed up issuance and cut manual error.
Yet, as adoption scales, so does scrutiny. Irdai’s FY25 report flagged mis-selling as a significant concern, and grievances rose 20 per cent to 257,000 complaints during the year, with claims related issues being the major concern. In this environment, deploying AI without embedding fairness, transparency and human oversight is not innovation; it is risk multiplication, the cost for which will be humongous.
I don’t see responsible AI as a compliance checkbox. It is the difference between a model that predicts risk accurately and one that inadvertently excludes a first-time buyer from a low-income household because the data it learned from ‘never included’ people like them. For an industry trying to extend cover to over 1.40 billion people, that distinction matters enormously.
Any honest conversation about responsible AI in Indian insurance today starts with the Digital Personal Data Protection Act, 2023. Insurers are among the most data-intensive businesses in financial services – collecting health declarations, clinical histories, financial records and behavioural data across a policyholder’s entire lifecycle. Under the Act, insurers act as “Data Fiduciaries”, obligated to obtain explicit, informed consent before processing personal data, and to justify why an AI system used that data the way it did. With fines up to Rs 250 crore for non-compliance, AI models used for risk scoring, pricing or fraud detection now needs an audit trail - not just accuracy, but explanation and consent hygiene built in from day one, not retrofitted after a regulator asks.
For me, four disciplines are non-negotiable as AI becomes embedded across the customer journey, and I think of them as inseparable from the technology itself, not layered on afterward.
Transparency In Decision-Making: Customers and regulators alike must be able to understand, in plain language, why an AI system approved, priced or declined a policy – not just receive an outcome.
Fairness By Design: Underwriting and pricing algorithms must be tested continuously for bias across geography, gender, and income and occupation. So, AI narrows India’s protection gap rather than widening it for underserved segments.
Data Privacy as a Foundation, Not an Afterthought: With sensitive health and financial data increasingly powering personalisation, insurers must build privacy safeguards into AI systems from day one, not retrofit them after regulatory prompts.
Human Oversight Where it Matters Most: Claims involving ambiguity, medical complexity or customer hardship still need a human in the loop. AI should accelerate decisions, not make the final call unsupervised on matters of life, death, and financial security.
Done right, responsible AI can be one of the levers for closing India’s insurance gap. It can bring underwriting turnaround times down from days to minutes, extend micro-insurance to gig workers and rural households through alternative data models, and use predictive analytics to nudge customers towards the right cover rather than the product with the highest commission. Irdai’s “Insurance for All by 2047” vision depends on this kind of responsible, inclusive scale.
Insurers who treat responsible AI as core strategy, not a side initiative, will be the ones who convert India’s underinsurance challenge into its next chapter of inclusive growth.
The next era of life insurance in India will not be defined by how much AI insurers deploy, but by how responsibly they deploy it. That responsibility is not a constraint on innovation; it is what will make innovation last for long.
The author is MD & CEO of Bharti AXA life Insurance
