Can insurance delivered over the blockchain be labelled trustless?

  • 30 May 2018

There is growing attention being given to how blockchains can help insurance. And amid the exciting claims being made for this development is one that warrants closer scrutiny. It's claimed that a blockchain can secure and record the transfer of any kind of digital data between multiple parties in a completely trustless manner. Yet what does trustless mean, and is this a realistic claim?

Some basics first. A blockchain is a continually growing list of records (called blocks) which are linked and secured using cryptography. Each block typically contains a cryptographic hash of the previous block, a timestamp and transaction data. Together they form a distributed ledger that can record transactions between parties in a verifiable and permanent way.

Beyond the State of Trust?

So why are blockchains being described as trustless? Trustless means no longer requiring the state of trust: in other words, beyond the need for trust. And it's a label being attached to blockchains through essentially a three sided argument. Firstly, blockchains are seen as replacing a centralised entity such as a bank with a decentralised, consensus based arrangement. In short, a ‘we don't need trust because we are all in this together’ type argument.

Secondly, rather than trusting in a mysterious, centralised entity such as a bank, blockchains are open to continual scrutiny by the distributed network of participants in them. And thirdly, when a transaction is entered onto the blockchain, any subsequent change to it will cause the verification that holds it there to fail.

What this certainly does not amount to is a technology that goes beyond the need to be trusted. Blockchains do not eliminate the need for trust. What they do is distribute trust across the many different actors in the system. This moves trust from being a one to one thing with say your bank, to a one to many, aggregated trust built upon the interests of multiple participants. What you have is a shared level of trust.

So how high or low is that level of trust? This depends on a mix of internal and external factors. Internal trust factors include the veracity of a blockchain’s underlying software, its governance structure and the rather big assumption that all participants in that blockchain share the same values. External trust factors include the completeness and accuracy of the information being put into the blockchain in the first place. None of those are free from influence or error.

A Mixed Reputational Record

The reputational track record of blockchains is mixed. They’ve been hacked and misused on several occasions. Taken together, this does not make blockchain into some form of magic wand. They do not make the data in them accurate or the people entering that data trustworthy. And while a consensus on trustworthiness across the participants in a particular blockchain has value, it does not eliminate the need for them to keep their critical faculties switched on.

Can you trust an insurance policy delivered via a smart contact on a blockchain more than one delivered by an insurance broker? You could do, but you shouldn't. It all depends on those multitude of factors that add up to make something trustworthy (more here). And that assessment should be pretty similar to how you weighed up the reputation of that insurance broker in the first place. 

There's a danger that as more insurance is delivered over the "trustless blockchain", people will take it for granted that the policy is just right for them, with all the cover they need and none of the exclusions they don't need. That would be a brave call for the insurance provider to make. 

Technology of any sort will never evolve to a stage that obviates the need for ethics or trust. To think otherwise is to misunderstand what technology is. Blockchain has much to offer to certain types of transactions in insurance, but it will not move any part of the market beyond the need for trust to be something earned and confirmed.