"Zurich is investigating applications of the technology such as extracting data from claims descriptions and other documents. It is feeding in the most recent six years of claims data in an attempt to identify the specific cause of loss across a whole section of claims, with the aim of improving its underwriting."

FT 24th March, 2023

ChatGPT has pros and cons which have been well documented.  See: -

Part 1: Chatpocalypse Now

Part 2- Chat-tastic Results

CIOs, CTOs, Underwriting and Claims Leaders must wonder of GenAI is a tool to deliver a competitive advantage and worry about being left behind. The key point is that you cannot hide away from it. In Stop Tinkering with AI I argued that insurers cannot just initiate isolated AI projects and expect to transform the business. They have to commit the right resources, leadership sponsorship, and capabilities to decide where to start and how to manage a technology "stairway to heaven" that leads to long-term competitive advantage. 

At the heart of the matter is data and the truth is that every insurer is an ocean of data islands all unconnected and inherited over decades of M&A and a major limitation to any AI-powered project. Lurking inside insurers are islands of AS400s and dotted around the ocean are some shiny digital islands trying to avoid being swamped by tsunamis of immature RPA, AI, and analytics projects. 

If you have not succeeded with RPA what makes you think ou will succeed with GenAI?

There are many open sources Gen AI tools and these need integration with scalable data management that can join ALL these incompatible data sources and build a data mesh and data fabric to share insight securely across the organisation.

That is a first step to experimenting with GenAI in which case consider: -

Companies SHOULD use these technologies with: -

  • Smallish, proprietary datasets
  • Well-bounded use cases
  • High-quality training data
  • Human expert oversight. 

Given these parameters, GAI can be extremely useful in AUGMENTING human workers, not replacing them.

To quote Geoffrey Moore who knows a thing or two about innovation, transformation, and technology "As we have noted elsewhere, digital transformation is not a restaurant. You cannot simply pay for it and have it delivered to your table. It is a gymnasium. You still must pay for it, but to get any value out, you have to actually do the transformational work yourself. Transformational software, in other words, is like a Peloton—it’s cool, but only if you engage."

Ericson Chan, who was poached in 2020 from Ping An to be Zurich’s chief information and digital officer, told the Financial Times that AI could create “a huge amount of efficiency” in jobs such as extracting information from long documents and writing code for statistical models.

It's not just about GenAI but also the strategic leveraging of AI across the business. This week The Economist says: 

"Process mining will help automate business long before chatbots do".

It continues "Running a big business is complicated—often mind-numbingly so. Seemingly straightforward processes such as taking an order and receiving the payment can take thousands of possible paths, for example if an extra credit-check is needed, delivery has to be confirmed or a follow-up invoice sent. Though often necessary, the rigmarole complicates life for companies and slows things down. The resulting inefficiencies can cost businesses eye-watering amounts—between 20% and 30% of annual revenue, according to one estimate."

Just think of the supply and repair chain networks of insurers- 20% to 30% waste time, materials and errors? A combination of process mining across whole claims journeys to the completion of repair and replacement is complicated involving so many parties. Effective collaboration, communication, and orchestration need technology platforms that enable process mining, and orchestration in a rapidly changing world e.g. in wich claims inflation has caught many a carrier out as can be seeb in valuations.

The older, complex platforms that take a year or more to upgrade hinder such initiatives. The modern, micro-services, API rich and cloud-native platforms like EIS, Genasys, RightIndem, ICE,  Instanda, Ignite are designed to enable agile transformation.

Able to leverage the wide range of AI-powered innovations from point solutions such as Tractable, Shift, CCC and so many others to the ell-hyped potential of GPT. 

Chan said Zurich was “very, very open” about its use of AI tools such as facial recognition and did not use micro-expression technology. But, in general, he warned about focusing only on the “pitfalls” of AI. “There are so many areas where there is so much benefit.”

Follow www.insurtechworld.org to help keep on top of the opportunities and challenges of AI technologies.

Further Reading

From Prediction to Transformation across Claims

The real next big thing in business automation

What P&C insurers can do as claims inflation pressures rise in Europe

Speed of settlement, customer satisfaction and AI