In the world of insurance, Swiss Re is a global leader. The company is a leading wholesale provider of reinsurance and insurance, with operations across the world. At Swiss Re, artificial intelligence is a key technology for core processes in both internal and external initiatives.
As the Data Science Lead at Swiss Re, Yannick Even is part of a team delivering end-to-end advanced analytics projects across the group, managing AI governance and compliance with Swiss Re’s standards for advanced analytics management and responsible AI.
Last week, Yannick sat down with Lisa Wardlaw, host of Insurance Unplugged, the expert.ai-sponsored podcast series, to talk about how AI is being utilized within the insurance industry, and specifically in corporate applications at one of the world’s largest insurance companies. This extensive conversation covers so many aspects of AI and insurance that you will want to listen firsthand to learn about:
- Trends impacting the sector: especially the impact of synthetic data and smart city digital twins
- New ways to look at risk and AI-specific risks that need to be mitigated
- Innovative initiatives at Swiss Re where AI has already delivered substantial value
In the meantime, when it comes to AI and insurance, here are Yannick’s main takeaways that every insurer should be thinking about today:
Think with a Scale Mindset
AI is starting to be used at scale to really solve tangible insurance challenges. But rather than quick prototype thinking, companies have to think about scale. When you think only in terms of what can happen quickly, this may be difficult to scale. Instead, when you think about a prototype with the end vision of a scalable solution, this offers a greater opportunity for success.
Start with People
Rather than technology leading the charge, it’s more about the people, about the culture, about the governance, about the data. Don’t spend all of your time looking at the fancy new technology; you’ll need to spend just as much time looking at the people and cultural aspects, getting the support from the top, getting the framework, upskilling your employees to make sure they understand the art of what’s possible. Inspiring your teams is just as valuable as understanding the latest tech.
Build AI and Data Literacy
Make sure your business IT tech people understand what it means to be a data-driven company, with data-driven processes and data-driven decisions. What are the new risks that could emerge from being data-driven from AI enabled, how should people think about leveraging data, managing data as corporate asset, what can AI do today? What can it do tomorrow? And again, the amount of focused, energy and money you spend on tech, you need to spend similar amounts on upskilling your people and upskilling your governance processes.
For more information on assuring AI success, get The Roadmap to NLP Success.
Target Key AI Use Cases
Make sure you’re using AI to support these functions:
One of the main benefits that AI addresses is to further automation, especially when you have repetitive knowledge, task or processes. In insurance, this applies especially to two areas: Classification: Often, the written submission, especially with unstructured data, can be now automated with AI. Claims: Here, you have a lot of paper, a lot of data points where AI can come in and further automation.
Enable decision making support
Given all of the data that insurers have access to, AI plays a key role in generating powerful insight to augment decision making. Insurers should be using AI to bring in all of your unstructured data so that it’s available for decision making processes.
Disruption and innovation
With the resources and processing power at our fingertips, insurers should be leveraging data and AI to disrupt outdated processes. For example, reinvent what protection you can offer and at the same time, bring more transparency and trust into the solution. A data-driven, AI-based approach can help you build a more granular, more personalized risk model that is focused on prevention.
Embrace Ecosystem Thinking
Remember, you are part of an ecosystem—don’t try to build all by yourself. There is better data outside that you could use—find a way to access it by providing value to your data partner. There is better tech and better cloud competition power out there. So again, find a way to provide value to your big tech and partner. There is value you can bring with your expertise to governments, to a regulator around responsible AI and different framework. So, find a way to bring this value so that these players want to open the doors and help you build something at scale.
Your customer is a customer of a full ecosystem. And when you work in an ecosystem, you also need to understand that yes, you have much more opportunity, but also you have to see the customer journey within this ecosystem and not just the few touch points that you have as an insurer. This is important to build a scalable solution.
Explore more “Build versus Buy” considerations in this white paper.
Balance Fast and Slow
As always, sometimes the early adopters are moving a bit too fast. The insurance industry is heavily regulated and this requires more work in the ecosystem; no one can fix it alone, you need to work with data providers, you need to work with government, you need to work with new partners to gather the data, build the powerful AI model, and then build integrated solutions within an ecosystem that is becoming more and more invaded. You will also need to upgrade your governance framework and your culture so that people not only know how to work with data and AI, but they also understand how to work with partners, which is very different from what insurance was done in the past.
Listen to the full episode on Spotify: Insurance Unplugged with Yannick Even