The rise of gen-AI in insurance: key tools and considerations for insurance

In recent months there has been much talk about generative artificial intelligence (gen-AI) such as ChatGPT. Its proponents claim that organisations and even individuals will have to use gen-AI tools to remain relevant and competitive—no matter which industry they’re in. This includes banking, financial services and, more specifically, the insurance sector.

In July 2023, MarketsandMarkets said in a report titled “Revolutionizing Risk: The Influence of Generative AI on the Insurance Industry” that: “According to a report by Enterprise Apps Today, the gen-AI in insurance market size is expected to be worth around $5,543.1 million by 2032. This growth will be driven by the increasing adoption of gen-AI by insurance companies to improve their operations and customer engagement.”

Gen-AI models such as ChatGPT, encompassing AI and machine learning (ML), are ushering in new opportunities for insurers across the value chain—reshaping insurance processes, customer experiences, and business models.

Ahead of Intelligent Insurer’s webinar on “The Rise of Gen-AI in Insurance: Vital Tools and Considerations for Insurers”, Madhur Naidu, Senior Technical Manager, Edgeverve (an Infosys company) talks about how he sees the technology benefiting insurers and topics that will be covered in this 75-minute webinar on September 7, 2023 such as:

● How to gain insights into the potential for improving risk assessment accuracy, enhancing underwriting efficiency and streamlining the overall insurance process; and

● Learning how this technology can generate new and creative solutions, helping insurers to drive cost reduction, improve the customer experience, optimise risk assessment, turn data into insights and accelerate decision-making. Meanwhile, here are his thoughts on gen-AI, and tips on the tools and considerations for insurance of this promising but potentially disruptive technology.

What percentage of insurers are adopting gen-AI technology, and why?

Everybody is curious to know what gen-AI can bring to their processes. Around 40 to 50 percent of insurers are putting in an effort to understand how they can adopt the technology across their enterprise. They want to do proofs of concept to see how gen-AI can help them to improve their processes and efficiencies.

The technology is relatively new, and people are still concerned about how they can use it in a positive way, and not in a way that starts to eliminate the “mundane workforce”. That said, the enterprise ecosystem needs to be ready. It comprises slow movers into new technology, so they need to be ready for change management, transforming the way their processes work across the enterprise.

What opportunities does gen-AI create for insurers across the value chain?

The value chain in the insurance industry starts from sales and distribution, marketing, policy management, claims management, audit and compliance, and finance and accounting, as well as product management. They need to manage the products they are selling.

Each function has an opportunity with gen-AI. For example, in marketing, they need to have a campaign for users across segments, and gen-AI can help with automation by using existing templates and campaign materials.

Once the marketing team inputs the requirements into the technology, it will create the whole campaign on its own, based on the public internet and on previous campaigns—taking them from a manual approach to a review-driven approach. It allows small changes to be made and reduces effort and time by 70 to 80 percent.

On the commercial side there’s underwriting. For a new business as of today, the underwriter has the tacit knowledge of the policy and comprehension of the risks. Let’s say I’m a trucking company with 500 trucks—what is the schedule for servicing? Today it’s all manual analysis, and much of the work is being done by junior underwriters to assemble all the information from the last three years, including new policies.

The underwriter needs to know all this information to rate and code the policy for a certain risk. An intelligent document processing solution with gen-AI can create a summary from all the data, from the broker, and give that snapshot based on the data-driven information available. They can see any exceptions. They can then analyse the risk and the premium.

For some complex policies it takes up to three to five days, but the time spent on this can be reduced by half. It reduces the process of collating the data as it can all be automated. If the answer is not there, gen-AI can create the answer and allow the underwriter to make a decision.

How can gen-AI reshape processes, customer experiences, and business models beyond traditional business process automation?

The current set of processes are more facilitated to human intelligence to service a case. The underwriter is using all the data provided to them, with the human in the centre, and based on the tacit knowledge, the human makes the decision.

The process will change with gen-AI as most of the decisions will be recommended by the technology—similar to the incoming submission. It will be based on the data provided, recommending humans to make decisions based on historical ones they made previously. The process is becoming more optimised, but the final decision is with the human.

The process is being optimised and becoming more efficient. The underwriter can process more cases in a reduced amount of time. As to the customer experience, the broker is getting output much faster with gen-AI than traditionally since the process is more optimised. The broker doesn’t have to wait five days. There must be a thought process on whether the brokers can approach underwriters in advance, or the same day, to give a quote.

At the moment there is much to-and-fro between brokers and underwriters. By reducing this, it becomes more seamless and frictionless. Gen-AI can help to improve customer experience. Customers will be happier as the process will be faster. The giving of quotes will be reduced to a matter of hours.

How can the technology improve risk assessment accuracy, enhance underwriting, and streamline the overall insurance process?

In the context of underwriting, the system can understand a specific business. With anything and everything that’s available on the internet across the last 50 years, the technology can create a synopsis and summary to analyse whether it is a positive or negative risk. For example, this could be about hurricanes in a particular area. Underwriters might not have this information when they quote for a commercial property, but with gen-AI it can be provided.

The technology can do the hard work and provide a risk assessment summary. The underwriter is doing this work on their own, or unable to do this type of assessment. The review process could be more thorough and complete with gen-AI. It has made underwriters’ work easier. The data drives the factors for a policy and the process is therefore streamlined. There is acceptance for using gen-AI for quoting, binding, and inspection.

How can insurers move from hype about gen-AI to working solutions?

Insurers must target specific functions, and properly qualify that a gen-AI approach will provide a viable solution for that challenge. For example, in claims—today there are many frauds that are occurring in claims and settlement. Gen-AI can do a proper audit to recommend whether to release claims funds or not.

With gen-AI, if we decide in commercial automotive insurance that it solves a problem, it can be trained to look at a region’s historical data where there is more crime. This will enable the giving of recommendations and permit a human audit—allowing auditors to choose where there is a higher probability of fraud in a particular area.

The technology can gather the information and decide whether the case is genuine or fraud. We can pre-empt the frauds that are happening and provide a solution. The idea is to identify the challenge, and then use the technology.

There is a lot of hype that gen-AI will do everything, but that’s not the case. All the information can be gathered from documents on the internet to decide whether it’s a positive or negative case. We use document processing features to feed this to the technology, asking it to give a recommendation.

Learn more and hear about some real-life examples of gen-AI being successfully used in the insurance industry, by booking your place at our webinar “ The rise of gen-AI in insurance: key tools and considerations for insurance”, to be held on September 7, 2023 at 4:00pm BST.