AI Is Here Podcast Series: How Zurich Insurance Is Leading the Insurance Industry with the Transformative Power of AI
This is the second in a series of blogs on the “AI is here.” podcasts. Each blog in the series will highlight the insights from a specific industry leader as they describe how their organization is deriving significant value from AI today.
In this edition of the “AI is here.” podcast, Dan Faggella, Founder and CEO of market research and publishing company Emerj, speaks with Gero Gunkel, Chief Operating Officer and Data Science Leader at Zurich Insurance Company, Ltd. Gunkel speaks about how the generative power of the latest AI models, and Natural Language Processing (NLP) in particular, is transforming the insurance industry and what he believes are the most significant ways in which AI will change what is possible for the insurance industry going forward.
According to Gunkel, text based use cases, which is the automated analysis and processing of text, is the most prominent use case among large insurance organizations today. This is both because these organizations have so much of this type of data, in terms of sheer volume, and because they can do so much to get value for the business with it.
For Gunkel, the important thing is that they use generative models that don’t just pull, for example a name or address from a document, but they can also write a short summary of the entire document, just like a human would.
According to Gunkel, there are three common use cases in insurance:
- Contract analysis – NLP is used to enhance the insurance industries core products, which are contracts. For example, as an insurer, they often do not have a lot of contact with their customers until there is a problem and the customer files a claim. It is very important for both sides to understand what is and is not covered. NLP enables them to analyze contract wordings to see what could be interpreted differently by them and their customers.
- Process Automation – An example of this would be where the AI reads inbound communications, such as a text, emails, or documents and processes them for human consumption.
- Analyst augmentation – in this case a knowledge worker could use AI to read a document and produce a summary of the document. In an industry with as much text based content as insurance, the ability to analyze multiple documents, produce summaries, and make connections between them is quite valuable
These are the use cases where they get documents from internal or external sources, which could be presentations, scanned documents, emails, and more. They then extract information from them automatically without requiring human workers, streamlining these manual or monotonous processes.
When asked why text based automation is so important to the insurance industry Gunkel identified two reasons:
- Because this industry has so much of it. Of the content they have, it is estimated that 80% or more is unstructured, such as images or text.
- They have seen better and more advanced technology, such as pre-trained transformer models, that provide insights out of the box. As a result, it has become much easier to use these technologies in business.
Gunkel went on to make some predictions about the future of AI. He made three predictions:
- He predicts that AI will not take human jobs. Instead, it will reduce the monotonous, repetitive jobs that some people do now and they will move to ones that require human interaction and communication
- When you have the ability for information to be generated, you get the ability to not just make processes more efficient, but to create new service models. Gunkel believes that AI will enable his industry to develop new, and better products for their customers.
- He sees there being a risk in the small number of big AI companies and that vendor dependency on them can be a challenge. Gunkel discusses how to mitigate this challenge.
AI is here. Discover how you can fundamentally transform what is possible for your organization with the power of AI in weeks, not years, with the industry’s most powerful, full stack deep learning platform while reducing the need for significant investment in infrastructure, personnel, and other resources.