Published in



Solving the double dipping challenge in the insurance industry with blockchain, confidential computing and AI

  • UPDATE 11/12/20: ClaimShare won the Insurtech challenge 2020 by R3 and B3i. More information can be found here
  • UPDATE 16/12/20: ClaimShare demo can be found on youtube, click here
  • UPDATE 02/03/21: You can find the official ClaimShare website here


IntellectEU launched its ClaimShare fraud detection platform for the insurance market to solve the duplicate claims problem (also called double-dipping) by using confidential computing, Corda blockchain and artificial intelligence.

ClaimShare separates claims in public and private data and uses fuzzy matching algorithms to identify suspicious claims, shared on the Corda ledger. Once claims are suspected of being fraudulent, we use Confidential computing to match the private data, confirming the fraud attempt before the second payout happens for the same claim.

The pilot focused on auto insurance in particular, but can easily be replicated for other insurance products.

Want to know more? Read on or email Chaim Finizola at

1. Introduction

Research by Insurance Europe and KPMG has pointed out that in 2017 alone, the total amount of fraudulent claims in Europe is estimated to be 13 billion EUR and that in Ireland, 4% of all claims are found to be fraudulent. The paper concluded that all industry players agree that exchanging data can further improve fraud detection.

Insurers are investing heavily in fraud prevention mechanisms, leveraging the newest technologies to perform deep analytics and identify patterns of fraudulent behavior. Unfortunately, they are limited to their internally-generated data for these purposes. There is no industry data sharing standard and there is no production-grade platform to facilitate industry-wide data sharing given the regulatory constraints of sharing sensitive, personal information.

While insurers would benefit by sharing all fraudulent claims, the lack of collaboration proves to be especially problematic when fraudsters book multiple claims for the same loss event at multiple insurers whilst the payout of one accident (even if insured at different insurers) by law can not surpass the total amount of the loss. This type of fraud is known as double-dipping in the insurance industry.

2. Problem

Double-dipping relates to filing a single claim with two different insurance companies. The payout of one event can not surpass the total amount of the loss. There are multiple situations in which this event takes place:

  • A single-car accident claim at multiple insurers
  • Driver’s own and the at-fault driver’s auto insurance
  • Multiple insurers for the same policyholder
  • Auto liability insurance & health insurance claims
  • Insurance for your belongings & your laptop separately insured
  • A single medical procedure claimed at two insurers
  • Double legal cost coverage
  • Disability insurance & unemployment benefits
  • Multiple travel insurances
  • Many others

Collaboration between the different involved parties regarding duplicate payouts for the same claim has always been very challenging. Some initiatives exist with a central party that collects all of the claims at all insurers and can be solicited when the suspected fraudulent activity took place.

The need for a third party that holds so much sensitive information, the need to always have the most up-to-date information and smooth claims query, and recent compliance and legal rulings put this model under pressure.

However, until today, there has not been an industry-wide platform that could solve this challenge. No solution exists that enables the secured data sharing between insurers on a need-to-know basis, whilst simultaneously guaranteeing business privacy and being GDPR compliant. There just hasn’t been a production-ready technology that could unlock this form of collaboration.

Anno 2020, with the recent technological advancements, this is about to change.

  • Corda as a distributed ledger has proven its DLT value underpinning multiple industries initiatives, confirming its scalability and privacy design
  • R3 launched their Conclave SDK for SGX’s confidential computing implementation
  • Open-source AI algorithms are continuously improving

We got our heads together and came up with a solution that promises to solve the double-dipping problem using these cutting-edge technologies.

3. Solution

ClaimShare offers a duplicate fraud claim verification solution across insurers, significantly decreasing the number of fraudulent claim payouts. By enabling insurers to put public claims data on the ClaimShare ledger after verification, other insurers can check if their claim has already been paid out at another insurer. This while also providing full business privacy, so that no competitive information is leaked.

In the event that a suspicious claim is detected, i.e. public data of the claim is similar to the public data of one or more claims stored on-chain, the private data is compared using confidential computing. If the private data matches as well, insurers can be sure that it is a fraudulent claim and a second payout to the end-user should not occur.

Matching of the claims happens in two steps:

  1. Public data of a new claim is matched against all verified claims, with state-of-the-art and proven matching algorithms.
  2. Final matching of suspicious claims is done with private multi-party computations by comparing the private data without revealing content.

For public data matching, we use multiple algorithms. These can be optimized and changed using artificial intelligence.

  • Exact matching — Used mostly for text and numeric fields, the data being compared has to be exactly equal to be marked
  • Fuzzy matching — Used on text fields, the data is compared given a similarity factor of percentage, and in cases where the level of similarity is greater than the factor, the field is marked
  • Difference matching — Used for numeric fields, given a factor of difference, in cases where the difference between two numbers is smaller than the factor, the fields are marked
  • Date matching — Used for date fields, given a threshold in hours, in cases where the difference between the dates is smaller than the threshold, the field is marked
  • Geo Scorer — Used for location coordinates, given a factor of the number of kilometers, in cases where the two locations are closer than the minimum distance, the field is marked

For private matching, we currently only use the exact matching of private data information.

4. Technical Implementation

ClaimShare is designed as a layer below insurers’ current back-office systems, limiting any required changes to the bare minimum. The API-based integration is done by using IntellectEU’s state-of-the-art integration product Catalyst Integration Solution, known for being the preferred integration product for the Contour trade finance network. Each insurer will have a node participating in the network as well as the network operator (Oracle) and the regulator (read-only). The solution can run both in the cloud as well as on-site using IntellectEU’s deployment and management tooling Catalyst Blockchain Platform.

There are various arguments why we are convinced that Corda is the best protocol choice for this particular use case:

  • Its sub transaction privacy using confidential identity
  • Its network model + governance model flexibility
  • The integration with the Conclave SDK
  • The orchestration of conclave workflow using Corda’s flow framework
  • The privacy model that enables multiple insurance products on the same Corda network
  • Its peer to peer attachment service
  • Its scalability, throughput, privacy, and support

5. Enhancements and Possibilities

We have proven the solution for auto insurance today. However, ClaimShare can be extended to additional insurance products including health, theft, and damage claims and used cross-insurance products.

  • The next ClaimShare release will also enable the P2P sharing of proof (pictures) supporting the claim using Corda’s attachment service. These pictures can also potentially be used for matching public data using AI.
  • Fraudulent claims can be flagged and committed to the blockchain to prevent additional fraud claims at other insurers
  • Uncovered fraudulent actions can be linked to other insurance policies linked to the same identity (within or between insurers)
  • Insurers can gather additional (external) verified fraud data to optimize their machine learning models to detect fraud
  • ClaimShare can help with the automation of payouts (starting with low-value claims)
  • ClaimShare can be linked to proof of insurance and verifiable credentials

6. Insurtech challenge

The ClaimShare idea originated as a result of the Insurtech challenge organized by R3 and B3i with the goal to come with a Corda-based solution that could help insurers collaborate in an effort to combat fraud.

The competition was divided into three parts, where we first had to be selected based on the value proposition of the solution, the market fit, and the pitch. This gave us the opportunity to look into the business problem, talk to insurance experts, and validate the idea before moving forward. In the second phase, we focused on the business model, the solution design, the go-to-market strategy, and showcased a clickable demo of ClaimShare, which gave us the chance to demo the application during R3’s yearly flagship event, Cordacon 2020. During the final phase, we were tasked with creating a working pilot of the solution, showcasing our development capabilities and proving our technical approach.

During this phase, our team of engineers (AI, Corda, confidential computing) delivered a fully working pilot in just five weeks, once again proving IntellectEU’s technical excellence and experience in the emerging technologies space. For the business case, we partnered with KPMG leveraging their insurance expertise on claim handling and fraud detection.

We are delighted to have participated in this challenge and would advise any other team thinking about participating to definitely do so in the future. We have received a great deal of business and technical feedback from R3’s team and partners and overall had a lot of fun working on ClaimShare.

Would you like to know more about the solution? Interested in watching our demo? Don’t hesitate to reach out to Chaim Finizola at

7. IntellectEU

IntellectEU is an international technology company focused on digital finance and emerging technologies. Since 2006, we have been focused on financial messaging and integration, being a SWIFT service partner. Building on our rich experience in financial services and technology, we founded our DLT and blockchain tribe in 2014. Always pioneering to serve our clients best, we became a founding member of Linux Foundation’s Hyperledger in 2016. Today, we are working with all leading blockchain providers (R3 Corda, DAML, Hyperledger, VMWare, IBM) in the banking, insurance, capital markets, and telco space. Apart from blockchain and DLT, we’re also venturing into artificial intelligence and quantum computing by educating our clients and building products with them.




Digital Finance & Emerging Technologies

Recommended from Medium

Has Cyber Risk Increased Due to the Rapid Rise in Telehealth Use?

Mainframe Ambassador Update

@OtIhersideMeta The second wave is coming!

{UPDATE} 妖怪タッピング Hack Free Resources Generator

Play Fair Cipher Encryption Using Python3

PIXBY: Frequently asked questions

{UPDATE} Epic TV Word Search - giant television wordsearch puzzle (ad-free) Hack Free Resources…

Hoo Officially Launched Hoo Smart Chain (HSC) Testnet

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Chaim Finizola

Chaim Finizola

Leading BDM in emerging markets at IntellectEU, a fintech specialized in emerging technologies. Contact:

More from Medium

AI based Solution for Enhancing Low Light and Dark Images for Field Assistance in AR

Energy Harvesting 101

Fighting climate change with satellite data — and the tools to deliver it

Workflow Webinar : VR Headset