Insurance Claims Automation

OCR, AI fraud detection on claims and claims analysis; no named model found.

OCR and document tabular

0named models

107candidates

8deployments

8in Indonesia

Named Models

The research names no model for this use case.

Where this comes from

Models the research names for this use case. Research, with sources on each model record.

Who Runs It

production, 5contract signed, 1pilot, 1planned, 1

Asuransi AstraCountry: IndonesiaOCR and AI for motor policy issuanceVendor: not statedproductionDate11 Sep 2026Policy issuance cut from 16 steps to 2; target 70 to 80% automatedEvidence: company statementConfidence: lowSource: investortrust.id

Prudential IndonesiaCountry: IndonesiaAI for customer profiling and claims analysis against policy termsVendor: not statedproductionDate6 Aug 2026Applied per Chief Digital and Technology OfficerEvidence: company statementConfidence: lowSource: keuangan.kontan.co.id

PT Zurich Asuransi Indonesia TbkCountry: IndonesiaAI motor claim survey: policyholders send their own photos and AI counts scratches and dents and checks coverage (product and vendor not named)Vendor: not statedproductionDate7 Mar 2024Faster process and better data accuracy (company statement, no figures)Evidence: company statementConfidence: lowSource: infobanknews.com

Tokio Marine IndonesiaCountry: IndonesiaAI fraud detection on motor and e-commerce shipment claims (model not stated)Vendor: not statedproductionDate13 Dec 2023No figures statedEvidence: company statementConfidence: lowSource: keuangan.kontan.co.id

Allianz IndonesiaCountry: IndonesiaAI for claims analysis, decisions and early fraud detection, developed from 2019 (vendor not named)Vendor: not statedproductionDate4 Dec 2020More than 7,500 claims a month analysed; fraud detection accuracy above 80 percent; twice the output of manual analysis (company claims)Evidence: company statementConfidence: lowSource: tirto.id

Oona Insurance IndonesiaCountry: IndonesiaCamCom image AI for car damage assessment in vehicle insurance inspection; severity and fraud features to be addedVendor: CamComcontract signedDate2 Feb 2024None statedEvidence: company statementConfidence: lowSource: investortrust.id

Asuransi AstraCountry: IndonesiaAI fraud detection for health claimsVendor: not statedpilotDate11 Sep 2026Described as developedEvidence: company statementConfidence: lowSource: investortrust.id

Asosiasi Asuransi Umum Indonesia (AAUI) with PT PindaiCountry: IndonesiaAI for claim condition checks and fraud mitigation in motor and health insurance, developed with an unnamed Chinese partnerVendor: PT Pindai with an unnamed partner from ChinaplannedDate30 Jul 2026Trial with insurers hoped for Q3 2026; no resultsEvidence: company statementConfidence: lowSource: investortrust.id

Where this comes from

Documented uses by named organisations. The evidence column says who made the claim.

Candidate Models

TabArena: All Tasks, no imputation, snapshot 2026-05-06, 59 entries

A general benchmark for this model class, not a test on this use case. Elo across tabular tasks, anchored RandomForest default = 1000; TabPFN-lab researchers are involved. Higher is better. Elo has no zero, so the axis starts at 1,200. Showing 10 of 13.

AutoGluon 1.5Amazon maintainer-run1,700

TabPFN-2.6Prior Labs maintainer-run1,626

RealTabPFN-2.5Prior Labs maintainer-run1,604

TabICL v2INRIA maintainer-run1,598

RealMLPINRIA maintainer-run1,514

TabDPTLayer 6 AI maintainer-run1,460

TabMYandex maintainer-run1,448

LightGBMMicrosoft AI maintainer-run1,435

CatBoostYandex maintainer-run1,418

XGBoostDMLC community maintainer-run1,376

1,2001,5001,800

XGBoostDMLC community, tabularReleasedAug 2026open weightshas scoreslast 18 months

LightGBMMicrosoft AI, tabularReleasedJul 2026open weightshas scoreslast 18 months

PaddleOCR-VL-1.6Baidu, OCR and documentReleasedJun 2026runs on edgehas scoreslast 18 months

TabPFN-2.6Prior Labs, tabularReleasedMar 2026open weightshas scoreslast 18 months

CatBoostYandex, tabularReleasedFeb 2026open weightshas scoreslast 18 months

GLM-OCRZ.ai (Zhipu), OCR and documentReleasedJan 2026open weightshas scoreslast 18 months

DeepSeek-OCR 2DeepSeek, OCR and documentReleasedJan 2026open weightshas scoreslast 18 months

AutoGluon 1.5Amazon, tabularReleased2026open weightshas scoreslast 18 months

TabICL v2INRIA, tabularReleased2026open weightshas scoreslast 18 months

RealTabPFN-2.5Prior Labs, tabularReleasedNov 2025open weightshas scoreslast 18 months

Showing 10 of 107. All models of this class

Where this comes from

Derived, not researched: models in the directory whose class matches a class this use case uses, excluding family records and models shut down or deprecated. Ranked by how many fit facts they meet, then by release date. Each fit fact is read from the model record.

Use Case

Industry: finance and insurance
Model classes: OCR and documenttabular
Benchmarks or datasets: not in the notes
Regulatory notes: Surat Edaran Menkominfo 9/2023 Etika Kecerdasan Artifisial, Kementerian Komunikasi dan Informatika (now Kementerian Komunikasi dan Digital)Insurance claims automation is an AI-based service run by an electronic system operator, and the circular asks such operators to build and use AI on the values of inclusivity, transparency, humanity and security.
UU 27/2022 Pelindungan Data Pribadi, DPR and President of the Republic of Indonesia (administered by Kementerian Komunikasi dan Digital)Insurance claims automation processes personal and financial data of customers, which the law treats as a specific category of personal data, so the controller needs a lawful basis, must protect the data and must honour data subject rights.
POJK 22/2023 Pelindungan Konsumen dan Masyarakat di Sektor Jasa Keuangan, Otoritas Jasa Keuangan (OJK)Insurance claims automation deals directly with financial consumers, so the provider's duties on fair treatment, clear information, data protection and complaint handling apply.
EU AI Act, Regulation (EU) 2024/1689 (international), European UnionInsurance claims automation decides access to an essential private service; the EU AI Act gives credit scoring as an example of a high-risk use, relevant when the system is offered in the EU market.

Where this comes from

From the industry tracks T08 to T10 of the research notes.

R&D Topics

Links to R&D topics are awaiting review.

Where this comes from

Which R&D topics would most change this use case. Drafted by Claude from the topic definitions and the use case; proposed until approved.

Incidents

Auto-Insurance Photo-Based Estimation Allegedly Gave Inaccurate Repair Prices Frequently2021-04-13, financialAuto-insurance companies' photo-based estimation of repair price was alleged by repair shop owners and industry groups as providing inaccurate estimates, causing damaged cars to stay in the shop longer.
State Farm Allegedly Discriminated against Black Customers in Claim Payout2021-06-12, discriminationState Farm's automated claims processing method was alleged in a class action lawsuit to have disproportionately against Black policyholders when paying out insurance claims.

Where this comes from

Documented failures, harms or misuse in this kind of use, Indonesia first, then the region, then well-known cases elsewhere (AI Incident Database, OECD AI incidents monitor, reputable news). Not a claim about any deployment listed here.

In Neighbouring Countries

Etiqa Insurance SingaporeSingapore, production, 2025-07-02AI straight-through processing of travel delay claims; pilot of vision LLMs for claims document review
AIA SingaporeSingapore, production, 2023-09-04Straight-through claims processing: 98 percent of claims auto-assessed, 60 percent paid within 24 hours with no human intervention

Where this comes from

Documented deployments of this use case in Malaysia, Singapore, Thailand, Vietnam and the Philippines, each with its source.

Opportunities

Opportunities are awaiting review.

Where this comes from

Gaps found by rule in what the directory holds for this use case. Each is a proposal for review.

Links

IndustryFinance29 use cases, 46 deployments

Use CaseInsurance Claims AutomationFinance

Modelsnone recorded

Who Runs It

Asuransi AstraPilot, 11 Sep 2026

Asuransi AstraIn production, 11 Sep 2026

Prudential IndonesiaIn production, 6 Aug 2026

Asosiasi Asuransi Umum Indonesia (AAUI) with PT PindaiPlanned, 30 Jul 2026+4 more

Graph

Deployments 8

Allianz IndonesiaAI for claims analysis, decisions and early fraud detection, developed from 2019 (vendor not named), Indonesia Asosiasi Asuransi Umum Indonesia (AAUI) with PT PindaiAI for claim condition checks and fraud mitigation in motor and health insurance, developed with an unnamed Chinese partner, Indonesia Asuransi AstraAI fraud detection for health claims, Indonesia Asuransi AstraOCR and AI for motor policy issuance, Indonesia Oona Insurance IndonesiaCamCom image AI for car damage assessment in vehicle insurance inspection; severity and fraud features to be added, Indonesia Prudential IndonesiaAI for customer profiling and claims analysis against policy terms, Indonesia PT Zurich Asuransi Indonesia TbkAI motor claim survey: policyholders send their own photos and AI counts scratches and dents and checks coverage (product and vendor not named), Indonesia Tokio Marine IndonesiaAI fraud detection on motor and e-commerce shipment claims (model not stated), Indonesia

Sources

3 Ways Insurers In Asia Are Embracing Generative AI oliverwyman.com https://www.oliverwyman.com/southeast-asia/our-expertise/insights/2023/nov/generative-ai-transform-insurance-asia.html
Asuransi Astra Targetkan 80% Proses Klaim Dilakukan Otomatis dengan AI investortrust.id https://www.investortrust.id/financial/115958/asuransi-astra-targetkan-80-proses-klaim-dilakukan-otomatis-dengan-ai
Prudential Indonesia Terapkan AI Mulai Sisi Operasional hingga Pengelolaan Risiko keuangan.kontan.co.id https://keuangan.kontan.co.id/news/prudential-indonesia-terapkan-ai-mulai-sisi-operasional-hingga-pengelolaan-risiko

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Insurance Claims Automation", https://aurion.semai.id/usecases/fin-insurance-claims/, records of October 2026.