The Story
Consultation on IRDAI's proposed Public Insurance Registry closes today, 30 September, a month after the regulator released the paper on 1 September.
The registry is proposed as a digital public infrastructure for insurance: population-scale, interoperable and non-exclusionary, described as an access layer connecting trusted information across the sector rather than a single central database holding every record. It follows a March proposal for a consent-driven infrastructure covering the full policy lifecycle, from issuance through claims, grievance redress and dispute resolution.
According to the paper, the registry would hold a tokenised identity for every policyholder, cross-insurer policy and claim references, consent records, grievance logs, intermediary conduct records and an anomaly-signal engine. It identifies use cases across eight stakeholder groups: the public and prospective policyholders, insurers, reinsurers, intermediaries, regulators, financial institutions, government departments and researchers.
One proposal is an Insurance Risk Score, a unified consent-based score synthesising insurance history held by the Insurance Information Bureau with credit and other data for underwriting.
IRDAI has indicated the registry could run through the IIB, an autonomous body established in Hyderabad in 2009, with plans to convert it into a not-for-profit company under the regulator operating under a separate framework.
Feedback was invited on data architecture, identity frameworks, data standards, privacy and consent safeguards, commercial confidentiality, governance and implementation readiness, through the IIB's PIR portal or by email.
The proposal remains at consultation stage, with no operational implementation date established. Separately, IRDAI's consultation on distribution economics, which triggered a 36 per cent fall in PB Fintech's shares last week, runs until 25 October.
Why It Matters
The registry has attracted a fraction of the attention given to IRDAI's distribution paper, and it is arguably the more structural of the two.
The distribution consultation caps commissions. It is painful, it is quantifiable, and the market priced it within hours. The registry does something less visible: it proposes to make insurance information portable and comparable across the industry, and asymmetry of exactly that information is what distribution businesses have been selling.
Consider what an insurance aggregator actually provides. A customer cannot readily see what cover they already hold, what it costs, what has been claimed, or how one policy compares with another. An intermediary can, and charges for the service. A registry that gives the policyholder a consistent view of their own records across insurers, with consent, removes a good deal of that gap.
The Insurance Risk Score proposal extends the same logic to the other side. A unified, consent-based score combining insurance history with credit data would let an insurer underwrite a stranger with less reliance on whoever introduced them.
This is the pattern India has already run through payments. UPI made the rails public, and value shifted away from whoever controlled access towards whoever built something worth using on top. The insurance equivalent would put the data layer in public infrastructure and leave intermediaries to compete on service rather than on what they alone can see.
The Strategic Read
Read the two consultations together and a consistent design becomes visible.
The distribution paper caps what may be paid for selling a policy. The registry proposal removes part of the reason selling commanded that payment in the first place. One attacks the price of intermediation, the other attacks its necessity.
For India's insurtech cohort that is the more serious of the two, even though it produced no share price reaction. Policybazaar, Turtlemint, RenewBuy, InsuranceDekho and Coverfox are distribution businesses. They compare products, match customers to policies and earn a share of premium for placing them. Their advantage is that the customer cannot easily do this alone. A registry that lets a policyholder see every policy they hold across insurers, with status, premiums and claims, erodes that advantage directly.
The anomaly-signal engine and intermediary conduct records point somewhere else again. A regulator that can see conduct patterns across intermediaries in one place does not need to wait for complaints to identify mis-selling. That is supervision by data rather than by grievance, and it lands on distributors before anyone else.
None of this is imminent. The paper is a consultation, there is no implementation date, and building population-scale infrastructure in insurance is considerably harder than in payments, because policy data is heterogeneous and insurers have a commercial interest in keeping their own records proprietary. Bima Sugam has been slower than announced for exactly those reasons.
But direction matters more than timing for anyone deciding what to build now. The layer being commoditised is the one most Indian insurtechs occupy.
For daily, sharp analysis of the biggest moves in the Indian business and startup ecosystem, follow StartupFox.
