In the second of a three-part series of articles on blockchain in the insurance claims context, our Head of Research and Development, Karim Derrick, explains how distributed ledger technology has been put into action.
In my previous article I looked at how blockchain is essentially a ledger that, by virtue of the way blocks of ledger entries are linked, has the valuable property of making the ledger largely immutable - changing any historical entry in the digital ledger will immediately break the hashes that bind the chain of blocks. It is this property that allows blockchain applications to be a substitute for the role of institutions providing trust. For example, banks are necessary to maintain ledgers of our spending to ensure the money that we have can only be spent once and that each and every transfer of money is properly accounted for. Currency blockchains, such as bitcoin, do away with the need for an intermediary: the blockchain has trust built into it and miners are incentivised to maintain its integrity.
The team has already got some experience of building blockchain applications in the education sector using the IBM-sponsored open-source blockchain platform, HyperLedger, and so it made sense to use this as the foundation for our exploration.
The first application that we developed looks at the issue of insurance fraud. It was developed with the hypothesis in mind that fraudsters often exploit the fact that insurers maintain independent databases and so operate it in the gaps between those databases. A claim registered with one insurer would be unknown to the rest of the market. With blockchain, there is scope at least for insurers to both have their cake and to eat it. Insurers can record all of their claims onto the blockchain as hashes. As we saw in the last article, hashes create codes that cannot be reversed. However, since the codes that they produce are always the same, it means that we can compare the hashes of personal data without having to reveal the original data. The same can be done with entire documents. So, if a new claim with a particular medical report is hashed and then compared to all of the other entries on the blockchain, we can find matches without exposing anything other than the matching data. Insurers can post their claims onto the blockchain knowing that they will not be exposing any of their data to their competitors unless the competitors have matching data.
The second application we have constructed explores a motor-focused value chain. The situation we wanted to look at is where we have one or more insured drivers involved in a collision where there is no dispute over liability and where there is no doubt over any one party’s indemnity. In this situation it should be possible for a claim to be processed, approved garages paid, approved car-hire providers paid and the claim to be closed without any intervention. The KLAiMChain application that we have built provides each insurer with a page onto which their insured can register their claims. Each claim is then written onto the KLAiMChain blockchain. If multiple parties are involved, the other parties are also able to register their claim, in each case registering the other drivers involved, and who, in their opinion, is liable. If all claims align, the system will go to the next stage. If liability is in dispute then it falls out of the system and is delivered to a claims handler. If liability is not in dispute then the system will check that all parties are indemnified as expected by sending requests to the insurers. If all insurers confirm indemnity then the system will trigger a communication to the insured’s asking that they go to the approved garages and credit hire companies. Once repairs have been made and credit hire invoices are raised onto KLAiMChain, the system triggers automatic payment.
Both applications exploit key properties of any blockchain application. Firstly, they are both network applications. They bring benefit by presenting a collective memory to multiple parties without compromising their roles as competing insurers. This is achieved in part because each insurer can trust the output of the system since no entry can be tampered with to the benefit of one party over another. Once a claim is written to KLAiMChain, there it will stay.
This article was first published by Claims Media on 10 September 2018.
Read part 1: Block by blockchain: what it all means
Read part 3: Block by blockchain: the future