AI in insurance: balancing great ambitions and ethical risks
#TechWatchbySeb Weekly Issue #21 - May 3rd, 2021
Hello my friends 🖐,
and welcome back to the Monday #TechWatchbySeb ☕️. This week we will travel across the Netherland 🇳🇱 and the U.K. 🇬🇧 after sharing insights on how Insurtechs are leveraging AI technologies.
After the major fundraising done by the French Insurtech Alan, which pushed the company to Unicorn status, I was curious to explore this sector further, and more specifically how AI is being applied.
I found this topic very interesting, as Artificial Intelligence is an incredible tool for insurers to accelerate and optimise their processes, including adapting pricing to specific users.
This last point is an important and sensitive one as automation could lead to unfair bias or discrimination. I am therefore keen to also find out how insurers plan to implement ethical guidelines in their use of AI technologies.
Let’s start, with some definitions
What is Insurtech?
Insurtech refers to the use of technology innovations designed to maximise savings and efficiency from the current insurance industry model. Insurtech is a combination of the words “insurance” and “technology,” inspired by the term fintech. (from Investopedia.com)
What is Artificial Intelligence?
Artificial intelligence (AI) is the ability of a computer or a robot controlled by a computer to do tasks that are usually done by humans because they require human intelligence and discernment. (from britannica.com)
By merging those 2 definitions, we could say that Insurtech using AI can be defined as tech companies developing AI systems to save time, reduce costs, improve customer experience and increase profitability in order to optimise time-consuming processes such as underwriting, claims management, fraud detection, customer service. Beyond savings in terms of time and costs, AI can also help reduce human errors. These errors can be common due to factors like changing regulation or the breadth of information to analyse for fraud prevention.
You can find more details in the article: Infographic — Artificial intelligence in the insurance sector.
What are the main use cases of AI in Insurance?
An insightful article by AI Multiple shows their are several use cases of AI in insurance, such as:
Claims processing
Appeals processing
Application processing
Insurance pricing
Document creation
Responding to customer queries
Claim fraud detection
Personalised services
Within all those activities, the main interests of using AI are:
Reduce mistakes: in preparing the registration, but also to better adapt the regulatory changes
Offer more personalisation: An Accenture study shares that 80% of insurance customers are looking for more personalised experiences.
Detect fraud: 29% of insurers have admitted to lying to their car insurance company to gain coverage in the US.
What are the most funded companies offering AI solutions in Insurance?
Shift Technology 🇫🇷 provides AI-driven decision automation and optimisation solutions for the global insurance industry. The company has raised a total of $100M.
Luko 🇫🇷 is a neo-insurance company that provides home insurance and security technology. It features an AI-powered sensor designed for monitoring and preventing home accidents. The company has raised a total of $72M.
Decibel 🇬🇧 is a real-time intelligence platform that enables businesses to measure and improve online customer experiences. The company has raised a total of $66M.
Streetbees 🇬🇧 is a human intelligence platform that collects and analyzes offline consumer behaviour. The company has raised a total of $64M.
Tractable 🇬🇧 is a software company that develops artificial intelligence for accident and disaster recovery. The company has raised a total of $60M.
What are the most active corporate VCs?
One aspect that I find particularly interesting is how corporate venture capital funds or CVCs are investing in this scene, compared to traditional VCs. Indeed, most AI applications require large volumes of data to generate successful results. Therefore, startups operating in this environment are most successful when working closely with large insurers:
Axa Venture Partners: The VC arm of AXA has raised $850M and already invested in 74 deals;
Allianz X: The growth investor unit of Allianz has raised $1Bn and already completed 36 deals;
Munich Re: Even if no amount of fundraising were released, they have completed 3 funds and invested in 35 companies.
Also, it is very interesting to see that US investors are quite active in financing European Insurtech companies. For instance, Accel, the US-based VC, has invested in 2 of the most funded companies above (Shift Technology and Decibel).
What is the place of Ethics in AI for Insurance?
Finally, I did some research on the place of ethics in an industry that generates profits based on its ability to evaluate risk. This evaluation has to be based on criteria that are clearly defined when calculations are done by humans (actuary), but what about defining limits when it comes to AI systems and Machine learning?
The Dutch Association of Insurers has published a framework to help insurers prioritise ethical measures when using AI. This paper,try to list the topics on which Insurer should put strong importance. I have highlighted what is behind the 5th bullet point on Diversity:
Human agency and oversight
Technical robustness and safety
Privacy and data governance
Transparency
Diversity, non-discrimination and fairness:
When violations of fundamental rights, including unjustified discriminatory bias, cannot be avoided or excluded in data-driven applications, insurers will not deploy an application.
When opting to use data-driven systems, insurers pay attention to diversity and inclusiveness, especially for people at risk of exclusion or disadvantage due to special needs and/or a disability
Societal and environmental well-being
Accountability
To conclude, I think this is definitely a market to watch as Europe already has a rich ecosystem of startups, corporates, and investors enabling the growth of AI in insurance. However, this growth will only be sustainable if strong codes of ethics are implemented in parallel.
#TECHWATCHBYSEB — WEEKLY ISSUE #21
🥇Messagebird🇳🇱 raised a €200M Series C and acquired 24sessions
The Amsterdam-based cloud communications platform just closed a successful Series C of €200M and also announced the acquisition of the US-based startup 24sessions for €600M.
🥈Darktrace 🇬🇧 made a successful IPO on the LSE
Announced months ago, the Cybersecurity company has managed a successful IPO on the London Stock Exchange. The company was valued $2.4Bn on the day of the listing.
Expectations were high for this IPO, especially after Deliveroo’s disappointing debut on the LSE.
🥉The EIB issue its first digital bonds using Ethereum technology
A major event happened last week in the crypto world, with the European Investment Bank issuing €100M bonds using the Ethereum Technology. A strong message sent by Europe using blockchain technology, while the BCE is working on a Euro-coin.
So, that’s it for this week, wishing you a great week ahead🖐
Stay safe ❤️