On Tuesday September 26, our special Roundcube Advisor Conny Dorrestijn moderated a session at Fintech Vortex as part of StartUp Fest week in the Netherlands. The session centered around the application of IoT and AI in insurance technology – a broad enough topic to fill a week of debate and dialogue, but here’s the drill down of the three main messages the Roving Roundcube Reporter took home.
There is almost no limit to what could be done technically
Insurtech SaaS platforms such as www.insureapp.com present on stage, are live today offering car insurance based on driver behaviour, using IoT to detect and coach and offer tailor fit proposals. The same is available for home insurance: IoT devices detect risk and often prevent things going badly wrong. All wearers of fitbit style devices appreciate the mountains of data that become available and that can be applied in health insurance scenarios: sleep patterns, food habits, physical exercise and more. Personal risk profiles for the insurer, prevention plans for the user, motivational programmes from employer to workforce. But it goes further, the UK’s K-Safe goes further, having started in response to the growing amount of cycling incidents (with in-helmet IoT devices that summon an ambulance if heart rate drops below a level and/or inactivity of the cyclist occurs), today they adapt their technology to provide intelligent theft and rider analytics to insurers. And soon, we might hear Sinatra’s hit “I’ve got you under my skin” become the company cheer for many general insurance providers when we – like our pets – accept chips to monitor us and keep us safe.
What would be possible for insurers embracing IoT and AI
At the core of both IoT and AI lies data, heaps of it, so the first boundary for deployment is the legal framework in which insurance companies operate. Data protection is a hot topic around the world and with the impeding GDPR in the European Community, we have a good idea of things to come into force in a wider arena: opt-in by individuals is the only safe way to go. This in turn implies that in order to get customer consent, one has to offer something truly relevant to that very user, the insurer’s end customer. At the same time benefits have to flow back to the insurance company too: either in the sense of increased STP, loyalty, lower cost, more customer insight and preferably all of those together.
This means that apart from great niche insurtechs who offer very tailor made solutions to – self segmented – target audiences such as pure play social insurer Bought by Many, large general insurers might have to think in a different direction. Explore the value of prevention and speedy recovery with the help of data more so than just how can I sell more policies at a better price? CED group, on stage at FinTech Vortex, explained that: “Using relevant data to deliver an optimal personal experience after a loss or any event leading to a claim does not only cut cost but leaves a lasting – and often carried on – positive imprint from an institution not often favourably looked upon”. These are areas where no insurance customer will ever repel against the use of their personal data as it offers instant gratification and benefit.
What should we do as a society with the infinite possibilities of IoT and AI
Finally much of the debate in The Hague centred on the ethics of IoT and AI. Where regulators try to ‘keep the beast fenced in”, we all know that technological advances simply cannot be stopped and, that undesirable and unexpected side effects do occur. Only recently a group of AI and robotics leaders – including Tesla’s Elon Musk and Google’s Mustafa Suleyman – appealed to the UN for a ban on ‘killer robots’. The military use of drones, AI, robotic technology, etc. they felt can only lead to a third arms’ race with unprecedented impact on society at large.
Cynics might find it strange that leaders who are up to their ears involved in the deployment of the very same technologies enter this debate. But similar debates take place in the insurance board rooms today. Will highly personalised health insurance policies deride solidarity, for that in fact is the basis of all insurance practices? Will companies with employee ‘fitness’ schemes – even if allowed by their workers – not entrench too deeply into the private lives of their staff if they get access to sleep and food patterns, not even speaking about when and where all of this takes place? We find it desirable if a parent can follow, say, the teenage daughter’s driving behaviour and her whereabouts late at night. But this is also highly culturally driven and almost a personal parental choice. To which degree will IoT devices in assisted living environments lead towards less personal attention for those who need it the most? And when do advice robos totally take over personal financial advice, to mention another ‘small’ topic?
Moral, cultural and ethical debates will have to be at the heart of the deployment of this technology if we as a society – including enterprise, government and citizens – are to benefit.
For now, let’s embrace the ‘could, would and should’ holistically. And also don’t let worry stop you from experimenting and exploring. Daniela Rus, director of MIT’s lab for robots and AI in the US recently stated in a Dutch newspaper article: “I do not believe in a moment where robots become smarter than people. It is easier to send a robot to Mars than to teach them to clear the table or fold your laundry.”
Robots compute better than humans but are not creative. Herein lays the message for insurers today: “stay creative and thou shall survive”.