Applying Blockchain & AI in Insurance

The insurance industry has been around for many centuries. While some history points to Chinese merchant seafarers pooling together their wares in collective funds, the first insurance contract stems from 1347 in Genoa. With that in mind, it is no wonder that traditional insurers have legacy systems and are slow to adapt to the changing world.

As the old adage goes, ‘if it ain’t broke don’t fix it’. It might be controversial to suggest the insurance industry is broken but it could certainly use substantial improvement. The last decade saw breakthroughs in technology where the world had more data than ever before, growing computer power and newly skilled roles like Data Scientists coming to the fore.

Two innovations have already started making their way into the insurance industry, showing themselves as highly beneficial. In this article, we are going to look at how artificial intelligence (AI) and Blockchain are being deployed by insurance firms to finally bring them forward into the 21st century.

What is AI? Blockchain?

Before we start telling you how they are disrupting the insurance industry, here’s a brief overview of what we mean when we talk about AI and Blockchain. Both terms have had so much hype recently; it is sometimes hard to cut through the myth and find the reality.

What is AI?

Although it is understandable to think that AI is a new-fangled term, its rise initially came about in the 1950s. Since then, breakthroughs like the Turing Test, which was the first time a machine exhibited human intelligence, have filled us with optimism for the field. However, on several occasions, with a lack of funding, skills and computing power, we saw “AI Winters” where progression stumbled and fell. 

In 2020, with vast amounts of data, technology, investments and skills available, as well as upcoming innovations including 5G, quantum and edge computing – AI is here to stay.

The world is in a time of what is called ‘narrow AI’. A machine with narrow AI capabilities is one that operates from a predefined set of rules. This could be the engine behind a Netflix recommendation or a voice command system like Alexa. Both are examples of artificial intelligence but are fed with criteria or training data in order to function. A good example is a driverless car which although impressive, is still narrow AI as it is given a set of rules to operate by. Until a car can understand the environment and think for itself, this will always be the case.

Narrow AI applications are driven by two subsets of AI known as machine learning and deep learning. The best way of explaining the link between the three is that AI is an all-encompassing term, inside of which is machine learning and then within this we have the more complex deep learning.

Machine learning applications are what we see being deployed in insurance. It’s the process of enabling machines to learn through data. The predictions the machine makes from that data is what we know as AI.

Let’s go back to the example of Alexa. Alexa receives a voice command, interprets it using an algorithm (known as natural language processing or NLP), matches the result against all existing data stored in the cloud to find the appropriate response, and sends this response back as a reply. Alexa gives the impression of being a cognitive machine but is far from it.

Insurance is ripe for machine learning given the vast volumes of data that companies have stored from years upon years of experience. We will look at some of its uses after we’ve given an overview of Blockchain.

What is Blockchain?

Blockchain is another technology set to transform insurance. The book “Blockchain Revolution” defines Blockchain as:

‘An incorruptible digital ledger of economic transactions that can be programmed to record not just financial transactions but virtually everything of value.’

It is a secure way of managing all transactions that happen in an organisation (or even industry if everybody obliges). As a digital ledger, a blockchain contains transactions that can be duplicated and distributed across an entire network of computer systems. Every individual block in the chain will contain a number of transactions and whenever a new transaction occurs, it is added to the ledger of all the participants.

A blockchain transaction is recorded with an immutable cryptographic signature known as a hash. This means that once data is verified, it cannot be deleted or edited. Hackers would need to corrupt every block in the chain across every version to where it has been distributed, making the whole chain incredibly secure. The graphic below shows the blockchain transaction process.

Source: Euromoney learning

Integrating Blockchain and AI

AI and Blockchain both have a lot of hype due to their disruptive nature. While they seem like completely different technologies, amalgamating the two together has the potential to revolutionise industries like insurance.

As we have discussed, AI involves using data for machine learning and vast amounts of it to make decisions. This means there must be a secure way of storing that data. As blockchain is decentralised, it can store transactional data on a network of computers that are continually validating the information, making it immutable (cannot be deleted or changed). Even if someone were to attempt to tamper with the data, the change (tampering) would be picked up by the other computers in the network and made invalid.

Data is stored cryptographically on a blockchain, making it perfect for storing the huge volume of personal information that AI needs to process in insurance.

Benefits of Blockchain technology

By incorporating blockchain technology into insurance, firms can expect several benefits:

  • Easy verification of transactions without having to rely on a third party
  • Data cannot be changed or deleted
  • Advanced cryptography to fully secure the data ledgers
  • All transactions are approved on a consensus basis
  • Distributed across every single node in the blockchain
  • It is decentralised, meaning there is no possibility of the data being lost across multiple machines
  • Transactions are transparent and open to anyone authorised to view them
  • Full audit trail for any set of transactions
  • Ability to pre-set conditions of a contract such as an insurance policy
  • Can virtually eliminate the potential for fraud or duplicate data entries

All these benefits can be applied to different insurance scenarios.

How Are Blockchain and AI Used in Insurance?

As we’ve already said, insurance has been around for a long time and whilst new technology has tried to infiltrate its way into the industry, there are still some companies lagging far behind. The number of online brokers and insurance companies has increased in the last decade but there are still cases where customers call in by phone and receive paper contracts, and pay through cheques.

With the amount of human intervention required in the insurance process, it is naturally prone to human error. In some scenarios, there are several stakeholders involved: customer, underwriter, broker, supplier, reinsurer and claims vendor, which could result in transactions getting lost or edited at any stage of the journey.

When it comes to AI, the insurance industry typically has a huge amount of processes. It is a sector full of documentation from policy wordings to insurance contracts, claims forms and brokerage deals. Using AI, these processes can be automated, improving productivity and reducing costs.

Below are some of the key applications of blockchain and AI in insurance. In some cases, you will see how they work well together while with others you will see how they are great as independent applications.

Fraud Detection

The Coalition Against Insurance Fraud estimates that $80 billion in fraudulent claims are made annually in the United States. In the context of insurance, a fraudulent claim is one where the claimant deliberately misrepresents the facts in order to gain benefits they are not entitled to. This could be through adding incorrect information or failing to disclose information, for example.

AI and Blockchain both offer solutions to help combat the problem of fraud. First, a blockchain network includes several insurers who could collaborate with each other and quickly identify suspicious behaviour. Instead of an insurer basing their decisions on transactions across their own data, they can utilise an immutable record across the industry.

However, there is a hurdle to overcome: it might take many insurers investing in blockchain before there are enough benefits aiding the detection of fraud. The most substantial benefits could come from the world of AI.

With the potential to process vast amounts of data in the cloud, AI applications such as machine learning could spot anomalies in real-time. Fraud alerts can be sent to relevant parties and malicious damages would be prevented proactively, rather than reactively, in response to potentially major incidents.

Large insurers like AXA are already using AI to combat fraud. They use a software called Darktrace, based in the UK, which troubleshoots threats related to highly skilled cybercriminals. In other words, complex algorithms spot anomalies in datasets within a split second.

Claims Processing

The claims process for insurance companies and their customers can be exceptionally long and laborious. There tends to be several parties involved, an often unnecessarily vast amount of documents required and, at the end of all that, no guarantee the claimant will be paid.

AI and Blockchain can potentially solve the problems associated with claims. For example, one emerging problem area post COVID-19 is claims from cancelled flights, which are often covered as part of travel insurance.

Let’s say a customer buys a travel insurance policy which includes coverage in the event of a flight being cancelled. This transaction enters the blockchain and it’s verified by the policyholder and the underwriter.

The insurance company then deploys algorithms which track airlines for flight cancellations. The data can feed straight back into the blockchain and map flight details with the customer’s information. In the event a cancellation is flagged, in theory the claim could be paid without the customer even needing to notify the insurer. We are not aware of any companies doing this yet in the travel industry, but it is a highly viable solution in a sector that is very volatile to high value claims.

Innovative property insurer Lemonade has reported that their system, which is founded on AI, can settle a claim in less than three seconds. It does this by assessing the information on a customer’s home and comparing it to their policy. A fraud algorithm then decides whether to accept or reject the claim.

This process works so well for Lemonade because it is intertwined with Blockchain. The business model of Lemonade takes a fixed fee from each monthly insurance premium and allocates the rest towards future claims. If a claim is made, the smart contract can verify it almost immediately against the terms of the digital contract and enable incredibly fast pay-outs.

AI techniques like computer vision can turn complex images into data, speeding up the claims process. For example, if a driver had an accident, they could take a photo of the damage and send that directly to the insurer. Computer vision will be able to interpret this image and make a decision as to whether the claim should be upheld or not.

Health Insurance

In the health insurance sector, systems require many back-office processes that validate whether it is the insurer or the patient who should pay for a service. However, with disparate systems and sometimes complex communication channels, there can be unnecessary delays caused by parties in the process. Patients get a poor experience.

For example, consider the number of stakeholders involved in obtaining authorisation from insurers. The hospital staff will need to speak to the patient and see whether they have relevant insurance coverage. The coverage will be investigated, to see if the procedure they are having is included within the underwriting terms. Next, the insurer will provide authorisation and potentially have to re-check on more than one occasion. This all comes before attempting to arrange payments and fees.

Deploying a smart contract as part of a blockchain could immediately check whether a procedure is covered. Within the contract could be all the terms from when the coverage was taken out, along with the authorisations needed to verify procedures. The health insurance blockchain provides the accountability and verification which doesn’t exist in most medical processes today.

Hospital staff can be relieved from using legacy communication channels and can instead focus on treating patients, while a blockchain takes care of the administrative tasks.

Proof of Ownership

The Economist estimates that the value of counterfeit goods sold worldwide each year is worth as much as $1.8 trillion. Product owners are faced with tremendous losses and desperately need some type of protection.

As we know, blockchain transactions cannot be edited or deleted. If an asset were to be listed or transacted on a blockchain, unless the owner verifies a change, the ownership is immutable.

In 2017, music streaming service Spotify teamed up with a New York-based blockchain start-up in order to register, identify and track creative works across the internet. Timestamps can be put on data to match royalties with the respective artist and mitigate disputes about the ownership of a track. Payments can be made on the basis of a smart contract, just like in other types of insurance. The move came after a high-profile case where Spotify had to pay out over $30 million to a publishing group over unpaid royalties.


Insurance underwriting can be a complicated process. If we take home insurance, an insurer needs to understand the state of the: property, surrounding area, previous claims history, customer details, mortgage information and more.

With AI this can all be automated and deployed to a blockchain. For example, there are several companies that provide property information such as BehindTheBricks or LexisNexis to name two. There is also direct integration with sites like Zoopla and Rightmove available if insurance companies have the skills in-house to develop their own connections.

When a customer begins a home insurance quote, an API maps this data to one of these providers, using a post/zip code and property number in most cases. All of the information about that property is retrieved without the customer needing to answer any questions.

The data can be stored against the customer record as part of the blockchain and form a smart contract. For example, if the property is found to have a large tree five metres away from the building, this can be stored in the blockchain. If the customer claims for a tree damaging the property, this can either be accepted or rejected instantly depending on the smart contract terms.

AI and blockchain can cooperate to create a fully automated underwriting process.

Big Data and the Internet of Things (IoT)

While both of these terms have been hyped for several years now, they have the potential to transform the insurance industry. The last decade has seen advances in computing power, cloud technology, skilled resources and volumes of data. Mainly, this has come from the exponential growth of digital channels as preferred ways for consumers to communicate and get information.

Chances are, anyone reading this article has some kind of mobile device, wearable technology, smart home sensor or voice-activated computer (Alexa/Google Home). All of these devices are incredibly useful day-to-day but the data they are collecting can be revolutionary to the insurance world.

Smart home technology is now a big business. Companies such as Neos are providing this innovative technology to customers as a way of reducing risk at home. The all-in-one insurance package includes devices such as cameras, smoke alarms and leak detectors, designed to collect data and protect the home. For example, a leak detector will send a mobile alert if abnormal water flow is found in a pipe. The alert might suggest a risk of a leak and notify the customer to check the pipe or switch off their water.

The objective here is to remove as much of the risk as possible associated with home insurance underwriting, meaning insurers won’t incur such heavy claim costs. As well, the additional data will allow insurers to personalise premiums and stop penalising customers who do take adequate precautions at home.

Data can play a vital role in many aspects of insurance. One example is Layr, which is a commercial insurance platform designed for smaller businesses. The machine learning algorithm which is built into the platform analyses vast amounts of customer data. In doing so, it is able to match an applicant to the policy that best suits them using a technique known as clustering.

Clustering represents a data algorithm that splits customers into segments automatically based on their respective attributes. Using these attributes, as well as predicting the right products for customers, it could forecast future risks or modify policy prices accordingly.

All verticals of insurance can benefit from IoT devices and data. In motor insurance, telematics devices can gather data as to how safely people drive. Insurers use this to price premiums more fairly. In health insurance, there is an opportunity to utilise the huge amount of data made available by wearables like fitness trackers. If an applicant is taking care of themselves with regular exercise, they could receive cheaper premiums.

Collective Health uses AI to remove the process of manually studying thousands of documents to understand the medical requirements of each of their customers. Machine learning integrates claims data, customer profiles and medical records to paint a holistic picture which helps to provide the right insurance coverage.

Documents and Natural Language Processing (NLP)

Smart contracts stored as transactions on a blockchain will allow insurance companies to either greatly reduce the need for documents or eradicate them entirely. For example, an insurance policy is often sent out with documents that can be tens or hundreds of pages long depending on the sector.

If each of the terms in the policy wording is specified and verified within a blockchain, there will be no need for lengthy paper documents. The result will be massive time savings and an environmental boon from printing less paper.

From an AI perspective, a machine learning technique known as natural language processing (NLP) can change the way documents are dealt with and processed.

Taking a standard insurer claim, there could be handwritten or anecdotal notes regarding the case. Insurers will likely have a team of people reading through the notes, entering them into a system and categorising them. NLP solves this problem for insurers.

Going back to Lemonade again, we established earlier in this article that they can settle a claim in three seconds. What’s more impressive about this fact, is it doesn’t even need a human involved. Conversational chatbots are digital or virtual agents which can understand human language and respond with intelligent answers. Alexa and Google Home are considered voice-activated chatbots.

These bots understand what we say to them and then use machine learning algorithms to translate texts or voices into data. The data can be mapped to a back-office data warehouse or blockchain, then verified, before a decision is made against the claim.

Computer Vision

Computer vision takes NLP to the next level. While NLP understands written or spoken words, computer vision technology intelligently interprets images (photos or videos).

Liberty Mutual is a great example of an insurer using computer vision to their advantage. In the event of a motor claim, users can take a picture of their damaged vehicle using a smartphone and submit it to the AI Auto Damage Estimator. The machine learning algorithm will match the images against thousands of others to assess the damage and the cost. The entire process takes only a few seconds without any forms.

As it stands, only a limited number of insurers are using computer vision. The reason is that the necessary computing power and connectivity have only been developed recently. However, as 5G starts to proliferate the market, it will enhance the ability for real-time interactions like we have seen with Liberty Mutual.


This article presented some of the ways in which blockchain and AI could transform the insurance industry. Many insurers are already adapting to the modern way of working but there is a long way to go. The limitations in insurance are not due to a lack of desire but to an industry that operates with legacy systems and extensive regulations. Deploying new, ‘black box’ solutions like AI can be challenging.

However, as AI and blockchain continue to permeate the industry, other companies are starting to jump on board. The number of blockchain networks is growing and insurers are finding a need to transform to keep up with the threat of insurtech and fintech businesses.

Disclaimer: The author of this text, Jean Chalopin, is a global business leader with a background encompassing banking, biotech and entertainment.  Mr. Chalopin is chairman of Deltec International Group,

The co-author of this text, Robin Trehan, has a bachelor’s degree in economics, a master’s in international business and finance, and an MBA in electronic business.  Mr. Trehan is a Senior VP at Deltec International Group,

The views, thoughts, and opinions expressed in this text are solely the views of the authors, and do not necessarily reflect those of Deltec International Group, its subsidiaries, and/or its employees.