After a six year pause, I have eventually returned to this blog. I wanted to write a series of posts providing advice to insurance leaders on how to successfully use AI in their companies. It is all my own opinions and based on my experience, which means that it is not meant to invalidate other people’s views and approaches. In fact, there are usually multiple ways to achieve similar results and I do not claim to know them all or to have the best one. This is just my contribution with the hope that you find it useful.

The first thing worth mentioning is that you do not have to be an expert in artificial intelligence to lead an insurance company that makes the most of AI. However, you do need to have access to such expertise and leverage it with good business acumen. Eventually you will learn what is possible, what is hard, and where your team’s expertise begins and ends. Very valuable knowledge.

In my opinion, what you need to successfully apply AI can be grouped into four areas (in no particular order):

• A business strategy with defined goals (measurable ones make it easier)
• A data infrastructure that is fit for purpose 
• A data science team
• A governance and ethical framework

I will cover each area in more detail in future posts.

Then it is your role to communicate and implement the company’s vision, and to embed a culture that facilitates responsible use of AI. 

However, the issue with not being an AI expert yourself is that you will have to trust your experts. That trust is very important and achieving it is not only your job, it is your data scientist’s job too. They must understand that it requires honesty and transparency on both sides. I plan to cover some aspects of work culture in later posts too but you should not underestimate their importance in the successful adoption of AI.