Three decades in this industry have taught me to think of trust not as a sentiment, but as a ledger. Every premium paid on time, every service request handled without friction, every small correction to personal details, each one is an entry. Reputation is the ledger. Economics is simply what that ledger produces, years later, when a family needs the promise honoured.
Most of these entries are quiet and easily forgotten. A customer may spend 20 or 30 years with an insurer and interact with it only occasionally: a premium reminder here, a policy service request there. Then comes the one moment that decides whether everything logged before it meant anything at all.
Why life insurance is not like other financial products
A home loan may run its course in seven years. A mutual fund can be redeemed on a bad Tuesday. Life insurance is different. It is often a 20- or 30-year relationship, taken well before one knows if or when it may be needed. Industry data bears this out: the Insurance Information Bureau’s analysis shows the share of life insurance policies with a sum assured of ₹2 lakh or less fell from 65% in FY21 to nearly 54% by FY25, with every higher cover bracket growing steadily through that period, which tells me something important: customers are backing an insurer’s ability to honour a promise decades away, not simply today’s premium calculation.
That confidence has to survive every touchpoint in between. A customer can move between an advisor, a branch, a website, an IVR call and a contact centre in the same week. What matters is whether they can pick up exactly where they left off, regardless of which one they used. It sounds like a simple ask. Delivering it consistently, across millions of customers and three decades of relationship, is one of the harder things an insurer has to do.
Building for real lives, not average ones
Customers do not live identical lives. A salaried professional, a self-employed entrepreneur, and a woman securing her family’s future think about protection differently, because their lives look different. The answer is not 500 products. It is building flexible enough products to respond to how people actually live.
This is visible across the industry. Term insurance uptake among self-employed individuals and women, particularly salaried women, has been rising. At the other end, India had an estimated 77 million adults with diabetes in 2019, a figure projected to reach 101 million by 2030, a population standard underwriting has historically priced out with loadings of 25-100% or outright declines; insurers are only now beginning to price this risk fairly instead of avoiding it.
Where AI genuinely helps
This is where AI changes the scale of what is possible. I expect that, in time, every process in insurance will carry some element of it.
Used well, AI can pick up life-stage signals, income shifts and changing needs in real time, often before a customer has articulated them. It can make conversations more contextual, sharpen underwriting summaries, strengthen servicing, and help frontline teams act with more precision. Across the industry, AI already supports sales pitches, coaching, customer profiling, and parts of underwriting and servicing. New policy administration systems are letting insurers configure and launch products in hours rather than months, something that would have been unthinkable a few years ago.
Transforming the value chain
The opportunity extends across the value chain. McKinsey identifies sales and distribution, underwriting, claims, servicing and back-office functions as major areas where AI can transform insurance. Its 2025 research also points to a shift from isolated pilots towards enterprise-wide AI transformation. Deloitte’s 2025 research similarly describes generative AI as having the potential to reshape how insurers operate, assess risk, launch products and interact with customers, while highlighting the readiness challenges that remain.
Governance has to come before speed
But adoption at scale brings a different responsibility. The regulator has taken note. IRDAI set up a dedicated AI working group in mid-2026 to build the sector’s first formal governance framework, prioritising explainability and human oversight in claims and underwriting. That mandate is playing catch-up: almost 82% of insurers already have machine learning integrated or in pilot, with predictive models influencing roughly three-quarters of underwriting decisions in life and health. That discipline matters because insurance handles some of the most comprehensive personal and financial information a customer shares.
The same thinking applies to speed. A one-minute policy approval is genuinely useful when the requirements are complete and clean. But a death claim may reasonably take three or four days, because that is what thorough verification requires. Medical claims can often be settled in minutes; other products may need longer. AI’s job is to make the right processes faster, not to make every process artificially instant.
What trust asks of technology
Explainability, to me, is becoming as important as accuracy. A customer should be able to ask where their data came from and get a clear answer. An employee or advisor should understand why a decision was reached, not simply accept that it was. Regulators will expect nothing less, and audit-readiness has to be built into the architecture from day one, not bolted on afterwards.
Technology can scale reach. It can improve productivity, personalise engagement and sharpen decision-making. What it cannot do is manufacture trust.
A product can be copied. An algorithm can be replicated. What cannot be copied is a decade of customers choosing, again and again, to trust their insurer with what matters most to them. That is why trust remains life insurance’s strongest moat. Increasingly, the winners in the industry will be defined not by who digitises fastest, but by who does it most responsibly.
(The author is MD & CEO, Bajaj Life Insurance. Views are personal.)