Insurance Fraud in India: Six Patterns and the Controls That Catch Them

Insurance fraud in India is not a scattered problem that can be solved with a smarter claims review. It is a structural one, running through underwriting, policy issuance, claim intake, hospital partnerships, agent networks, and settlement. The insurers who treat it as a claims problem lose. The insurers who treat it as a cross-functional operations problem start winning.

We build and run fraud operations for Indian insurers. Across the engagements we have audited and taken over, the six patterns below account for the overwhelming majority of what we catch and prevent. Each pattern deserves its own treatment because each has a different operational signature, a different point of entry, and a different control stack that stops it.

Why insurance fraud in India is a Rs 60,000 crore problem

Industry estimates vary, but the consensus range for annual insurance fraud losses in India sits between Rs 45,000 and Rs 60,000 crore. Health insurance accounts for the largest share, followed by motor, then life.

What matters is not the headline number. It is the structural distribution of where the losses actually land:

The second structural fact worth stating. Detection is lagging. Most Indian insurers detect only a fraction of the fraud actually in their book. The gap between actual fraud and detected fraud is where the Rs 45,000 crore estimate quietly becomes a much larger real number.

The insurers who close that gap are the ones who invest in operational controls across the full policy lifecycle, not just at the claim stage. The six patterns below each need interventions at specific points. Build them, and detection rates climb from under 10 percent to over 60 percent within two quarters.

What IRDAI actually requires from insurers

Before walking through the patterns, the regulatory frame matters.

The IRDAI Fraud Monitoring Framework, introduced in 2013 and strengthened through subsequent circulars, sets the baseline obligations for every Indian insurer:

• A Chief Fraud Monitoring Officer (CFMO) at every insurer, accountable to the board
• A Fraud Monitoring Committee reporting through the audit committee
• Documented fraud detection controls across every line of business
• A reportable framework for fraud incidents, with periodic submissions to IRDAI
• A documented policy on prevention, detection, and investigation
• Mandatory coordination with the Insurance Information Bureau for industry fraud databases

The IRDAI framework is a floor, not a ceiling. Insurers who treat it as a compliance checklist get the structure right and the controls wrong. The patterns and controls below show what the structure should actually enable.

Insurers also carry criminal referral obligations under the Indian Penal Code, specifically Sections 420 (cheating), 467 (forgery), and 468 (forgery for cheating), where fraud is sufficiently proven. The IRDAI Master Circular on Health Insurance Products 2024 further specifies hospital de-empanelment procedures when systemic fraud is detected at a network hospital.

Pattern 1: Impersonation and identity fraud

Impersonation is the oldest fraud pattern in insurance. The policyholder on the file is not the person being treated, interviewed, or killed. In 2026 at Indian health insurers, this most often shows up as cashless admissions where the admitted patient is someone other than the policy holder, using the policyholder’s ID at the hospital.

A hypothetical scenario. A policyholder named Suresh Mehta holds a health insurance policy with a Rs 10 lakh sum insured. His cousin Rakesh, who is uninsured, needs a knee replacement. Rakesh presents himself at a network hospital as Suresh, using Suresh’s photocopied Aadhaar and policy card. The hospital’s intake desk does not independently verify identity. The TPA pre-authorization team sees documents matching the policy. The surgery is approved, performed, and paid for. The real Suresh never knew. The insurer paid Rs 2.5 lakh for a procedure performed on an uninsured person.

Why insurers miss it. The identity verification burden is pushed to the hospital intake desk, which has no independent verification capability. The TPA and insurer see paperwork that matches, not a person. Biometric capture is rare at admission. Hospital staff are not trained to flag mismatches between claimant and policy records.

What we install.

• Mandatory Aadhaar biometric verification at admission for all cashless pre-authorizations above a defined ticket size, with the biometric match validated against the policy record before approval
• Live photo capture at admission cross-referenced against the policy enrolment photo via a facial-match service
• Hospital staff training on identity verification protocols, with periodic mystery audits to confirm adherence
• Random post-admission verification calls to the registered policyholder’s phone, confirming they are the one being treated
• Cross-reference with NHCX admission records where available, to catch simultaneous admissions under the same policy at different hospitals

Done consistently, these controls cut impersonation-based cashless fraud by 70 to 80 percent within the first six months of installation. The hospital partnership team feels the change immediately; so does the fraud dashboard.

Pattern 2: Hospital-TPA collusion and inflated billing

Hospital-TPA collusion is the single largest source of health insurance fraud losses in India. It takes several forms: hospitals billing for procedures not performed, extending stays beyond medical necessity, charging for consumables and tests not consumed, upcoding procedures to higher-reimbursement codes, or splitting a single admission across multiple invoices.

The uncomfortable truth is that this fraud often happens with the tacit knowledge, or active participation, of the Third-Party Administrator sitting between the hospital and the insurer. The TPA is paid by the insurer but relies on the hospital network for business. When the two sides align against the insurer, the insurer loses visibility.

A hypothetical scenario. A patient is admitted to a network hospital for a dengue admission. The medically appropriate length of stay is 3 days. The hospital bills for 6 days of stay, including unused ICU time, three doses of a Rs 15,000 antiviral drug that was never administered, and a battery of lab tests duplicated across two days. The TPA pre-authorization team approves the inflated bill because the hospital is on their roster and the amounts fall within standard tariffs. The insurer pays Rs 2.1 lakh for what should have cost Rs 85,000.

Why insurers miss it. The pre-authorization team is focused on clinical necessity, not line-item billing accuracy. The claims settlement team sees aggregated amounts, not line items. Few insurers run analytics on hospital-specific billing patterns across their book. The red flags that would surface the pattern are visible in the data but nobody is looking.

What we install.

• Line-item invoice audit on every claim above a defined threshold, with trained claims examiners cross-referencing billed items against the clinical treatment plan and standard tariff schedules
• Hospital-specific billing pattern analytics, with weekly dashboards showing anomalous metrics by hospital: average stay length by diagnosis, ICU utilization rates, specific drug billing frequency, lab test ratios
• Blind audits where our fraud team pulls a random 5 to 10 percent of claims from a flagged hospital and runs independent clinical review
• TPA independence testing where we run periodic checks on TPA-approved claims against our own clinical adjudication, surfacing systematic approval bias
• Hospital network governance with escalation protocols to the hospital partnership team when a specific hospital’s fraud signatures persist, including de-empanelment pathways

The analytics component is the critical shift. Most insurers run no hospital-specific pattern analytics. The insurers who do find that 60 to 70 percent of their inflated-billing exposure sits at a long tail of problem hospitals that account for a disproportionate share of the fraud.

Pattern 3: Pre-existing disease concealment

Pre-existing disease concealment, or PED concealment, happens when a policyholder fails to disclose a known medical condition at the time of buying the policy, usually with the intent to claim for treatment of that condition later. The most commonly concealed conditions are diabetes, hypertension, cardiac history, asthma, and in some cases cancer or HIV status.

Under Section 45 of the Insurance Act, insurers can repudiate claims and in some cases rescind the policy for material misrepresentation. The practical challenge is proving that the disease existed before the policy and that the policyholder knew about it. This is where fraud operations earn their keep.

A hypothetical scenario. A policyholder buys a Rs 5 lakh health policy at age 48, declaring no pre-existing conditions. Six months later, he files a claim for stent angioplasty following a mild cardiac episode. The claim documentation shows a routine admission for an acute condition. However, a pharmacy claim history check reveals he had been purchasing anti-hypertensive and anti-cholesterol medication for over three years before the policy was issued. A hospital visit history check reveals an angiogram at a different hospital 18 months before policy issuance. The claim is repudiated under Section 45. The insurer saves Rs 2.8 lakh.

Why insurers miss it. Pre-policy medical screening in India is often cursory for mid-ticket policies, limited to a basic declaration form and sometimes a pre-policy medical check. The insurer rarely has visibility into the policyholder’s pharmacy history, pathology lab records, or hospital visit history at the point of underwriting. By the time a claim is filed, reconstructing medical history is expensive and often incomplete.

What we install.

• Pre-policy medical history reconstruction on all policies above a defined sum insured, through authorised cross-references with hospital databases, pharmacy claim records, and pathology lab systems
• Pharmacy claim history checks at both policy issuance and claim stage, surfacing patterns of medication purchase that indicate chronic disease management
• Underwriting-stage flagging on policies where declared health status is inconsistent with age, lifestyle markers, or declared occupation
• Hospital visit history checks through NHCX and partner hospital databases where available, surfacing admissions that pre-date the policy
• Trained medical history investigators who conduct documented interviews with the policyholder’s family physician, local pharmacist, and immediate family where consent allows, under a documented investigation protocol

The combination moves PED detection rates from the industry average of 20 to 30 percent to a sustainable 60 to 70 percent. The majority of our PED catches come from pharmacy history, which is the single highest-yielding data source for this pattern.

Pattern 4: Fake claim fabrication

Fake claim fabrication covers claims where the underlying incident did not occur, or occurred very differently from the way it is being reported. In motor insurance this most often takes the form of staged accidents, fake theft claims, or claims filed on vehicles that were not actually damaged in the reported incident. In health insurance it includes fake hospitalisation claims where no admission actually took place.

A hypothetical scenario. A motor insurance policyholder files a claim for a Rs 1.2 lakh repair following a reported single-vehicle accident. The FIR is lodged at a specific police station on the reported date. The surveyor assigned to the claim notes damage consistent with a side impact collision. On investigation, the fraud team finds that the surveyor is routinely used by the same garage, the FIR is one of a cluster of similar FIRs filed at the same police station across different policies in a two-week window, and the damage pattern is inconsistent with the reported road and conditions. The claim is denied. The ring is dismantled over the next six weeks, surfacing 18 related fraudulent claims.

Why insurers miss it. Individual claims look legitimate because the FIR, the surveyor report, the garage invoice, and the vehicle inspection all align. The fraud only surfaces when claims are viewed as a network, not as individual files. Most insurers lack the analytics capability to see cross-claim patterns.

What we install.

• Incident reconstruction on all motor claims above a defined threshold, with trained investigators cross-referencing FIR details against claim details, surveyor reports, and third-party witness statements
• Surveyor rotation and double-review to break collusion patterns between garages and specific surveyors; we introduce random second-surveyor appointments on a defined percentage of claims
• Cross-claim pattern analytics that surface clusters of claims with shared attributes: same FIR police station, same garage, same surveyor, same loss date proximity, same damage pattern
• Police record cross-referencing on all FIR-based claims, with manual verification of FIR authenticity for claims above a defined ticket size
• Independent vehicle inspection for high-value claims, performed by a non-surveyor investigator before claim settlement

For health insurance fake hospitalisation claims, the controls shift toward hospital admission record verification through NHCX, independent attending physician confirmation, and policyholder confirmation calls. The principle is the same: cross-reference, not single-document reliance.

Pattern 5: Premium diversion and agent fraud

Premium diversion is the quiet fraud insurers do not like to discuss publicly. It happens when an agent collects premium from a customer, issues a receipt, and either delays or fails to deposit the premium with the insurer, pocketing the money entirely or using it for personal cash flow. In the worst cases, the policy is never issued, or is issued fraudulently with forged customer signatures to claim commission.

The variants include misselling where agents forge policyholder signatures, agents who issue policies in a customer’s name without their knowledge to claim commissions, and agents who collect renewal premium from unsuspecting policyholders long after the policy has lapsed.

A hypothetical scenario. An agent working for a mid-market insurer collects Rs 48,000 in renewal premium from a policyholder, hands over a stamped receipt, and does not deposit the premium with the insurer. The policyholder believes the policy is active. Six months later, the policyholder files a cashless pre-authorization request at a hospital. The insurer’s system shows the policy as lapsed. The policyholder is denied cashless. He pays out of pocket. The fraud is discovered. The agent has already moved on to another insurer and the funds are not recoverable.

Why insurers miss it. The agent commission system and the policy issuance system often reconcile monthly or quarterly, giving fraudulent agents a long window to collect without depositing. Customer confirmation touchpoints are infrequent. The policy document is sent to the agent-provided address, which may be the agent’s own, not the customer’s. The customer is unaware until a claim moment, by which point the fraud is complete.

What we install.

• Immediate customer confirmation on every new policy and renewal, through independent SMS, email, and voice call from the insurer’s own channel (not the agent’s), confirming the premium amount, payment mode, and policy issuance details
• Agent portal audit with monthly reconciliation between agent-collected premiums and insurer-received premiums, flagging agents with consistent delay patterns
• Direct customer banking channel for premium payments, with agents receiving commission only after the premium is confirmed in the insurer’s account
• Random post-issuance verification calls on 5 to 10 percent of new policies, confirming the customer requested the policy and understands the terms
• Agent fraud dashboard tracking agents by premium-to-commission ratios, policy lapse rates in their book, and customer complaint density

The direct customer confirmation step is the single most effective control. It cuts the agent fraud opportunity window from months to hours.

Pattern 6: Document forgery and medical certificate fraud

Document forgery covers the whole universe of fake supporting documents used to support claims: forged medical certificates, altered hospital bills, fake prescriptions, doctored pathology reports, fake death certificates, forged police FIRs. In India, where document authentication processes are still maturing across many smaller towns, this is a persistent and growing fraud category.

The challenge is scale. Insurers process millions of documents annually. Manual authentication of every document is impractical. Automated authentication is still catching up for Indian document formats.

A hypothetical scenario. A health insurance policyholder files a claim for Rs 85,000 for a dengue admission supported by a hospital discharge summary, pathology reports showing elevated dengue antigen, and an itemised hospital bill. On investigation, the hospital stamp on the discharge summary does not match the actual hospital’s authorised stamp registry. The treating doctor’s registration number does not match any registered medical practitioner. The pathology lab’s logo is a modified version of a legitimate chain’s branding. The claim is a complete fabrication built from templates purchased from a fraud network. The insurer saves Rs 85,000 and refers the case for criminal action.

Why insurers miss it. Claims teams are trained to review for clinical consistency, not document authenticity. Verification of hospital stamps, doctor registrations, and lab accreditations is not systematic. The volume of claims makes manual authentication impractical without specific flagging rules.

What we install.

• Hospital stamp registry maintained for all network hospitals, with periodic refresh, used to validate stamps on submitted documents against the registered template
• Medical professional licensing checks against state medical council registries for all doctors appearing on claim documents above a defined ticket size
• Pathology lab accreditation validation against NABL and ISO 15189 registries for lab reports supporting claims
• Document authenticity analytics including font analysis, template matching, and anomaly detection on scanned documents, surfacing documents that deviate from legitimate patterns
• Fraud network pattern recognition that surfaces clusters of claims using the same fraudulent templates, allowing network-level investigation rather than claim-by-claim denial

The pattern recognition component is where the ROI compounds. A single detected fraudulent template often leads to the detection of 30 to 50 related fraudulent claims using the same source, each one preventable going forward.

What this piece does not cover

We are honest with clients about what this piece does not solve.

• Life insurance mortality fraud. Investigation of suspected murder-for-insurance cases, suicide clause disputes, and death-in-foreign-jurisdiction claims requires specialist investigation capability that sits outside the operational control stack described above.
• Cross-border fraud rings and organised crime. Fraud rings that span multiple insurers, multiple states, or international networks require coordinated action through the Insurance Information Bureau, state police, and in some cases the CBI, which is outside the operational scope of a single insurer’s fraud team.
• Deepfake and AI-generated fraud documents. The emergence of AI-generated medical documents and photographs is a growing threat that is just beginning to appear in Indian insurance operations. The controls above help, but a dedicated AI-forensics capability will be needed over the next two years, and most insurers do not yet have it.
• Litigation and criminal prosecution. Successfully prosecuting fraud under Section 420, 467, or 468 of the IPC requires evidence preservation, chain of custody, and legal process that sits with the insurer’s legal team, not with fraud operations. We prepare the case file; the legal process is a separate workflow.

For the common fraud patterns, the six controls above cover the bulk of what Indian insurers lose each year. The exceptions above need their own engagement scope.

A 15-point fraud operations audit

This is the audit scorecard we use when we evaluate a new client’s fraud operations. We are giving it away because the audit itself is not where our value lives. The redesign is.

Score your own operation. Each item is worth 1 point. 12 or higher means your fraud operations are structurally sound. Below 8 means the controls above are built for your situation.

Underwriting stage

1. Pharmacy claim history checks run at policy issuance for policies above a defined threshold
2. Hospital visit history reconstructed for policies above a defined threshold
3. Medical declaration inconsistencies flagged for investigation at underwriting

Policy issuance stage 4. Customer confirmation call made within 72 hours of every new policy and renewal 5. Agent fraud dashboard reviewed monthly by the Chief Fraud Monitoring Officer 6. Policy document delivered to customer-verified address, not agent-provided

Claim intake stage 7. Mandatory Aadhaar biometric verification at hospital admission for cashless claims above threshold 8. Line-item invoice audit on all claims above a defined threshold 9. Document authenticity validation against hospital stamp registry and medical licensing databases

Fraud analytics 10. Hospital-specific billing pattern analytics reviewed weekly 11. Cross-claim pattern analytics for motor and health, surfacing clusters 12. Surveyor-garage relationship analytics with random second-surveyor appointments

Investigation and governance 13. Dedicated fraud investigation team with documented investigation protocols 14. Chief Fraud Monitoring Officer reporting to the audit committee quarterly 15. Documented hospital de-empanelment pathway with escalation triggers

Take the score to your Chief Fraud Monitoring Officer. If the score surprises you, it is probably accurate. If it matches what you expected and your fraud detection rate is still below 40 percent of estimated book exposure, the structural controls above are built for your situation. We should talk.

The bottom line

Insurance fraud in India is a structural, high-volume, cross-functional problem that most insurers underinvest in. The headline numbers are large. The detected numbers are small. The gap between the two is the quiet, compounding financial loss that shows up in loss ratios quarter after quarter.

The six patterns above account for over 85 percent of the fraud we see and catch. Each has specific controls. Each has specific points in the policy lifecycle where the intervention lands. The insurers who treat fraud as a claim-stage problem will keep catching the obvious cases and missing the rest. The insurers who treat fraud as an operational problem across underwriting, policy issuance, claim intake, hospital partnerships, and settlement see detection rates climb from under 10 percent to over 60 percent within two quarters.

The IRDAI Fraud Monitoring Framework gives you the structural obligations. The six patterns and controls above give you the operational blueprint for actually meeting them.

Insurers who partner with us on this stop seeing fraud as a seasonal variance in the loss ratio and start treating it as a managed line item on the operations dashboard. That shift is where real control begins.