Filip Lundin
Insurance companies: how to accelerate AI development
AI and machine learning are on everyone's lips. But despite the obvious potential and rapid technological development, the insurance industry is lagging. Filip Lundin, business area manager for insurance at Billogram, writes this and shares three insights on how insurance companies can position themselves at the forefront of AI development.
I've had the privilege of attending several AI events recently. During Insurance Evolution's network meeting a couple of weeks ago, two things in particular became clear: 1. There is enormous potential regarding the use of AI in the insurance industry, and 2. the insurance industry is falling behind.
With seemingly endless possibilities with AI and machine learning, it's therefore important to start at the right end and ensure that the focus is on creating concrete business value. So, how do you succeed?

1. Create a raw list of use cases
First and foremost, it's about identifying concrete challenges and problems in the business where AI can be useful. The key here is neither to get too technical nor to start from what the technology can achieve. Instead, companies must start at the other end—in the business—and involve all parts of the organization to create a raw list of concrete cases to evaluate and test. Johan Larsson from Mavera recently highlighted in a column that personal injury claims handling is an area where AI offers great opportunities for innovation and efficiency. Other common and well-defined challenges that can be quickly improved with the help of AI include, for example, matching incorrect payments and answering common customer questions based on the company's FAQ.
2. Organize the company correctly internally
Once business problems have been identified, it's important to create a dedicated and business-oriented team with direct exposure to management, which will work to bring the entire organization along on the change journey. Instead of an AI team deep within the IT organization or a new CX department, the team must be centrally placed, be given a clear internal change mandate, and high trust when it comes to testing, sifting, and learning—to be able to develop an understanding of where AI can generate the greatest business benefit.
3. Work with partners
Last but not least, there are great advantages to working with partners who have already tested, made mistakes, iterated, and found smart ways to use the technology to solve common use cases. With the right partner, the insurance company gets the opportunity to avoid others' mistakes—and more quickly copy successes.
Is AI mostly a gimmick, or can significant business value be created already today, the skeptic might wonder? If you look at other industries, there are many concrete success stories to be inspired by. One example in the telecom sector is Hallon, which has optimized its processes with the help of machine learning and now individualizes its invoicing and reminder flows. Through experimentation with AI, they have managed to halve customer churn within the group of late-paying customers—and at the same time significantly reduced the number of inquiries to customer service.
In other words, AI can already today create concrete business value and at the same time improve the customer experience. The conclusion is therefore simple. The technology is here and the potential is enormous—it is now high time for insurance companies to accelerate their AI journey!
For those who want to demystify AI and understand its basics, I recommend the University of Helsinki's open course Elements of AI (2 ECTS), which more than a million global students have completed.
Filip Lundin