Star Health bets big on AI: targeting 50% cashless claims through automation within two years
Star Health and Allied Insurance Company plans to lift AI-settled cashless claims from about 20% today to up to 50% over the next two years. The company is backing this with higher tech spend-rising from around ₹120 crore this fiscal to about ₹200 crore in FY27-spanning customer acquisition, claims, and cybersecurity.
As leadership put it, "Future is digital, tech, AI and we are investing heavily." The direction is clear: automate what's predictable, keep humans on complicated cases, and compress turnaround times without letting fraud slip through.
The numbers that matter
- AI in claims: ~20% of cashless claims are settled via AI today; goal is up to 50% in two years.
- Volume and speed: ~5,800 claims handled daily; 96% of cashless claims processed within three hours.
- Fraud baseline: Industry fraud is estimated at 7-8% of total claims.
- Data advantage: AI models trained on 20 years of data-market-level patterns, hospital length of stay, and disease cost curves.
- Growth & scale: FY25 GWP at ₹17,553 crore.
- Regional footprint (AP & Telangana): 24 lakh lives covered; ₹1,990 crore GWP till February this fiscal; 19% YoY growth.
Why this matters for insurance leaders
- Turnaround time: AI-first adjudication trims human touch on standard claims, improving cashless discharge and provider satisfaction.
- Fraud control: ML models can pre-score claims, triage to SIU, and flag anomalies by hospital, procedure, cost, and stay length.
- Consistency: Policy interpretation and package pricing become more uniform when models align with rule engines.
- Scalability: Higher straight-through processing (STP) supports growth without linearly adding headcount.
Where AI fits-and where it doesn't
AI can handle the bulk of pattern-based claims-routine procedures, clean documentation, and known cost bands. Complicated, ambiguous, or high-severity cases still need human judgment and, ideally, AI-assisted review rather than full automation.
Execution playbook for health insurers
- Data foundations: Standardize claim, pre-auth, and provider package data. Close gaps in ICD/ procedure coding and discharge summaries.
- Model strategy: Pair business rules with ML. Use risk scores to auto-approve low-risk claims and route medium/high-risk to human review.
- SIU integration: Feed flagged claims to investigators with clear reasons (outlier cost, LOS variance, duplicate billing, provider pattern).
- Human-in-the-loop: Set thresholds for auto-pay, assisted review, and hard holds. Capture reviewer decisions to keep training data fresh.
- Provider connectivity: Tighten hospital integrations for real-time documents and coding accuracy to cut back-and-forth and reduce discharge delays.
- Operational SLOs: Track STP rate, approval TAT, FN/FP on fraud flags, grievance rate post-settlement, and re-opened claims.
- Governance & explainability: Maintain audit trails of model inputs/outputs and approvals; ensure model changes are versioned and reviewable.
- Cybersecurity: Treat claims and health data as high-risk assets. Apply least-privilege access, encryption, and continuous monitoring across the stack.
Regional lens: AP & Telangana
The company covers over 24 lakh lives in Andhra Pradesh and Telangana, with ₹1,990 crore in GWP till February this fiscal and 19% growth over last year. Expect AI-driven STP and fraud scoring to focus on provider networks and procedures that dominate volumes in these markets.
What to watch next
- STP growth vs. customer experience: keep disclosures clear and escalation simple.
- Explainable AI for contentious claims to reduce disputes and regulator queries.
- Evolving guidance from regulators on health claims processing and grievance redressal. See the IRDAI for updates.
Resources
- AI for Insurance - frameworks, use cases, and tooling for claims, fraud, and underwriting.
- AI Learning Path for Medical Billers - workflows that align payer-provider claims operations.
Your membership also unlocks:
