How AI will revolutionise the insurance industry

html on screen

Many believe the “technological singularity”, or simply the “singularity”, is fast approaching. This is the moment when machines will match, and subsequently exceed human intelligence. What does this mean for the insurance industry?

Job losses versus improved services

In January 2017, Fukoku Life reduced its payment assessment department by a third after introducing a new AI system. Even before the redundancies, the AI system was being used to analyse customer voices, to deal with complaints, and generally improving the efficiency and accuracy of its operations. The Japanese life insurance company expects AI to increase overall business efficiency by 30% and yield a ROI within two years.

The machine at the core of Fukoku Life’s AI operations is IBM Watson.

IBM prefers to refer to it as a “cognitive” machine rather than “artificial”. Such machines are able to learn, reason, and interact naturally with humans, developing their abilities from their interactions with their environment. They are probabilistic, in that they generate hypotheses rather than simply numerically accurate answers. They are also able to handle vast quantities of unstructured data and keep pace with the complexity, unpredictability and increasing volume of information.

Swiss Re is also using IBM Watson in its Life and Health Reinsurance department. The reinsurance wholesale provider is using the technology to help insurance professionals make better informed insurance decisions, and to price risk more accurately.

AI and vehicle repair costs

AI can accurately estimate vehicle repair costs by analysing images of the damage in a process that takes just seconds.

The system is already revolutionising claims handling, carrying out the process in a fraction of the time and better than can be achieved by a human. It can also spot potentially fraudulent claims and flag them for investigation. Tractable, the developer of the system, claims that it isn’t designed to replace humans, simply to make humans better and more efficient at what they do, leading to a reduction in the cost of claims and improved customer outcomes.

Peer-to-peer insurance

Possibly the most disruptive threat to the traditional insurance business is peer-to-peer insurance.

Pioneered by the New York-based Lemonade, its technology uses an AI bot called Maya to create insurance policies. Lemonade claims it takes just 90 seconds to get insured and just three minutes to get paid. Its primary insurance products are renters and home-owners.

The start-up’s business model is very different from that of a conventional insurance company. Rather than pay out claims from the company’s funds as do conventional insurance companies, it pays out from a shared peer-to-peer resource and takes just a flat fee for providing the service. Any unclaimed funds are donated to charity. This means that Lemonade has no interest in withholding payments. To reduce the risk of fraud, Lemonade uses a range of sophisticated anti-fraud algorithms. Although there are other insurance organisations that adopt a peer-to-peer model, Lemonade is the first fully regulated peer-to-peer insurer.

The future of the insurance industry

It’s clear AI is having a huge impact on the efficiency and accuracy of processing insurance applications and claims. It’s also creating significant disruption to jobs within the sector. So what of the future; where will the insurance industry be a decade away from now?

One possibility is that more peer-to-peer insurance companies similar to Lemonade will emerge. The peer-to-peer model is already revolutionising financial services such as lending and investment; there appears to be no reason to suppose that it will have any lesser an impact on the insurance sector. Driven by technology, particularly in terms of intelligent anti-fraud systems, this could prove transformative for the big insurers. Perhaps, as is happening already in the financial sector, the big insurers will have no choice but to also adopt similar peer-to-peer models.

Some sectors are predicting that it won’t be long before we see a completely AI-based insurer. The bottom line is that most tasks in the insurance industry value chain have the potential to be improved by AI. It seems that intelligent machines may work alongside human workers – or replace them. Whether the future is evolutionary or revolutionary, the insurance industry will evolve as a very different business from what we see today.