By FintechOS · October 19, 2021
7 minute read

Rise of the machines: how AI is driving innovative insurance


Key takeaways

  • 84% of executives believe they need AI to achieve their growth objectives
  • To compete, you need an AI-ready platform

To learn more about our Northstar insurance platform, book a demo.

According to Finance Digest, artificial intelligence (AI) will empower 95% of customer interactions by 2025. Meanwhile, Accenture found that 84% of executives believe they need AI to achieve their growth objectives.

Those are difficult statistics to argue with. So, what makes AI so vital to financial services?

The key is data. AI feeds on it. The more you feed it, the more it can do for you.

The insurance industry, between the all-important Know Your Customer (KYC) data and risk information, has vast stores of data that can empower an AI. As such, the insurance industry is in a prime position to benefit from everything AI can provide.

What exactly is the benefit of AI, however, and what can it do for you that is so indispensable? Let’s look at what exactly AI is, what it is certainly not, and what it can do for your company.

What AI is and what it isn’t…

Doctor Hannah Fry of University College London took to Twitter a couple of years ago to address the state of the AI industry:

If data is the new gold, we’re living in the wild west. We’re surrounded by extraordinary claims, often with little proof that the algorithms can or will do what they say. Some work but don’t work well. Others are nothing more than snake oil.

Dr Fry quoted the case of an algorithm that promised to predict the sexuality of a person based only on their gender with an 81% success rate. The problem is, as Dr Fry noted, that 94% of adult men identify as straight, so simply saying all men are straight is more accurate than the algorithm.

There are a lot of AI tools out there that are promising to do a lot of things. Yet…

AI is not magic. It can’t yet do things that human beings can’t do. AI can, however, do some things that humans can do, but faster and more efficiently.

AIs don’t sleep, they don’t need breaks or holidays, and they can run many operations at once.

Say you’re considering a collaboration with another insurer. You need to find out how many of your life insurance customers have motor insurance with your partner company.

In the past, that might have been a six-month research project for a team of your analysts, or an expensive customer survey run through an agency. An AI program, however, can be repurposed to find the information you need from public records overnight.

That’s just the start. AI can be used to take on almost every fiddly piece of time-draining admin that absorbs your resource. Of course…

There’s fear that [AI] will leave people out of a job. Yet, this isn’t the case. Rather, AI frees up your people’s time to do the things AI can’t: creative thinking and innovation.

AI creators make some very lofty claims as to what AI can achieve, such as defining your strategy or predicting the future. AI is better suited, however, for doing the day-to-day while you focus on the important things you never have time for.

Far from robot overlords, AIs in action are more like the automated checkouts in a store being overseen by a single, human supervisor, doing the simple tasks to reduce the resource needed. Just like those automated checkouts, you don’t need to look to science fiction to see AI operating effectively in the real world.

Let’s look at some examples of how AI is benefitting insurance companies right now:

Tractable: motor claims damage assessment from a photo

London company Tractable have an AI product for the motor insurance market that can assess the damage from a motor accident from just a photo. This remote estimation lets you create accurate repair reserves, but also offers instant answers for your customers.

Customers involved in a motor accident simply take photos of the damage to their insured vehicle and upload it to the insurance company’s app. Tractable’s AI will then compare the photos to millions of other visual examples of vehicle damage and then accurately predict how much repairs will cost.

This gives the insurer an accurate reserve for the claim. It also offers an instant indication to the customer as to whether their vehicle will be repairable, a total loss, or cheaper to take to the garage themselves.

Of course, a skilled claims agent may well be able to make precisely the same judgement. The AI, however, will make decisions consistently across all claims without the variability or risk of human error. Not to mention, the AI will use the data from every claim made through your firm to improve its estimations.

The system is already being used by Admiral and Geico.

Fukoku Life: accurate automated medical insurance payments

The industry standard for AI is IBM’s Watson application. In 2017, Japanese life insurer Fukoku Mutual Life started using Watson to assess its claim pay-outs.

The system scans medical records and bills, claims history, and the policy to calculate the payment the claimant is entitled to. Claims handlers are only then required to authorize the amount to confirm the AI’s judgement.

Allowing Watson to do the claims estimation increased Fukoku’s productivity by around 30%. Not to mention, faster time to payment increased customer satisfaction.

Aetna: AI chatbots for customer satisfaction

US company Aetna was one of the first to adopt AI chatbot technology as far back as 2010. The company’s virtual assistant, “Ann”, has become a mainstay of their website, dealing with claim queries, estimates, and verifying identification.

The AI chatbot deals with 20,000 queries per day, and responds by text or voice. Starting with a 35% approval rating, Ann was rated positively by 80% of customers within a year. This is the real talent of AI: continual learning and improvement.

Like most chatbots, Ann can refer customers to a claims agent when it can’t deal with a query. This means simple questions can be dealt with immediately, even at busy times, and Ann can act as a queuing system when she’s not able to help, keeping traffic to the Aetna call center manageable.

Lemonade: claims payment in three minutes

Lemonade is the insurance success story of the last decade. The firm saw a nearly 30% increase in revenue in 2020 vs 2019, particularly driven by global lockdowns.

Lemonade doesn’t have call centers or claims handlers. Everything is processed through their app.

An AI engine called Jim can provide a quote for cover in under 90 seconds, while onboarding takes just minutes. Recently, a customer achieved a “new world record” for time to payment, receiving USD 729 in three seconds.

This all leaves customers happy and Lemonade focused on growing their company and innovating on their service.

FintechOS: bringing your AI together

Of course, to truly take advantage of all this, you need to organize all of your systems and workflows around a platform that can support them.

You need a platform that can connect your legacy systems to the sophisticated AI tools that will make your insurance services stand out from your competitors.

The FintechOS Northstar insurance platform comes with a range of integrated AI tools, including optical character recognition (OCR), facial recognition, and liveness detection. Not to mention, it can be easily connected to any third-party AI service or app.

Northstar can be implemented on-premise or to the cloud using a low-code install. This means you can be live with an AI-ready system in weeks, not months.

To learn more about our Northstar insurance platform, book a demo.

Share this article with your connections