Log in

Forgot password?

Don't have an account with us?

Register Now

Generative AI in Practice – Issue 1

Generative AI in Practice is InsTech’s monthly newsletter dedicated to the use of generative AI in insurance – you can sign up for free here.

The InsTech Perspective

Spotting the GenAI use cases that bring competitive advantage

The hype around the use of Generative AI is matched only by the FUD (Fear, Uncertainty and Doubt). Some forward-thinking insurance companies tell us they have identified between a dozen and a hundred possible use cases of generative AI. Some of the areas where generative AI can add value as a productivity tool are becoming clear, from analysing emails, data extraction and claims automation to coding and policy information chatbots. Certain insurers are revealing how they expect to use generative AI to gain competitive advantages, as the CEO of Travelers says he is looking to achieve (see below). But concerns about regulation, biases, hallucinations and customer privacy are causing some insurers to ban ChatGPT and other generative AI tools from the workspace. 

If you’ve read this far chances are you are curious about what Generative AI means for you and your business. Following the success of our parametric and climate newsletters, we’re testing out this pilot for a newsletter designed to help you identify those areas where generative AI can bring the most value to your organisation right now and understand more about where your peers see the opportunities. We plan to bring you news and insights every month. If you have a story you’d like us to feature or questions to answer please contact us at [email protected].

Use case of the month

Extracting information from emails

Large language models (LLMs) are one of the building blocks of Generative AI. LLMs are capable of summarising information and extracting data from documents of different formats. Several tools are now available to summarise long email conversations or converse with documents.

Many insurance companies receive submissions, claims and all sorts of other documents through an email inbox. The world would be a better place if insurers, brokers and their clients could exchange information through widely accepted portals rather than relying on the 21st-century version of a postcard. But change is never easy and so far now, the cutting edge of technology is being used to extract meaningful data from emails. The process of directing messages to the appropriate team and then extracting key information from emails and attachments into internal underwriting or claims systems is one of the use cases of generative AI where we have seen the most interest and some adoption from insurers and technology companies so far.

Guest Perspective

Generali: 3 factors to consider when adopting generative AI

Emanuele Colonella, Innovation Lead at Generali, is optimistic about several ways insurers can use generative AI. He sets out the factors insurers should be mindful of as they start to use the technology.

What the insurance industry is doing with GenAI

AXA deploys a generative AI service, developed with Microsoft, to 1,000 of its employees. The service, which it plans to roll out to its whole workforce eventually, can generate, summarise, translate and correct text, images and coding.

Travelers’ CEO says the company is looking to create “meaningful, sustainable competitive advantages” with generative AI and will spend more than $1.5 billion USD on technology in 2023.

MAPFRE’s research into the risks of AI will focus on operational risk, ethical risk and regulatory risk, while the company claims it is currently testing more than 80 use cases of generative AI.

Singlife collaborates with Microsoft on an accelerator to help insurance start-ups integrate generative AI into their products.

The Hartford bans ChatGPT while exploring 24 use cases of LLMs.

Indian insurer HDFC ERGO has chosen Google Cloud as its partner as it launches an initiative to build new GenAI-powered products and services.

Life insurance broker Modern Life has launched a generative AI tool to help life insurance advisors with policy information and client communication.

News from the InsTech network

The future of underwriting is the topic of InsTech’s next evening event in London, sponsored by CytoraJoin us on 19 September at CodeNode to learn about how underwriters can benefit from generative AI and other technologies.

Send Technology launches its GenAI-powered Smart Submission solution for data extraction. Listen to our recent podcast with CEO Andy Moss to learn more.

Inversion’s Founder Jacques Bosch explains how generative AI and semantic search can compare two versions of a reinsurance treaty or respond to customer service requests.

Eigen Technologies describes how intelligent document processing, powered by technologies such as LLMs, can help insurers deliver better services and generate more revenue.

A study from Future Processing found that 85% of UK consumers thought insurers were not adequately considering AI to enhance their customer experience.

Developments in generative AI will lead to an increase in the frequency of cyber attacks, according to CyberCube.

A blog post from Marsh explains that while the prevalence of generative AI could affect risks which are already insured, such as those related to data privacy or IP infringement, it can also create new risks.

When is the right time to create new insurance products for AI risks? George Beattie, Head of Innovation at CFCshares his thoughts.

Other news and insights we’ve seen

How generative AI’s ability to create deep fake images and videos will affect cyber liability insurance.

Even at insurance companies that have banned generative AI, employees are secretly using ChatGPT to get ahead with work, according to Business Insider.

How InsTech has used AI this month

This month, we’ve been testing a tool that writes meeting notes automatically, powered by OpenAI’s GPT-4. 10 seconds after a call, this tool sends us a summary of the call alongside a list of action items. One of our team members said this was saving them time on every call and was the best transcription tool they had seen for understanding technical insurance terminology, which can otherwise be hard to follow and summarise when taking notes manually. Rumour has it that our chairman has privately licensed an advanced version that ensures he has no actions assigned to him from the call. We’ll reveal our findings in our next release.


Generative AI in Practice is InsTech’s monthly newsletter dedicated to the use of generative AI in insurance – you can sign up for free here.

Stay up to date with the latest from InsTech