What Should Insurers Know About Generative AI?

In data-heavy industries like insurance, new digital tools are always making headlines. Whether it’s Robotic Process Automation, fraud detection, or workflow automation, there’s always something new promising sweeping change well into the next decade. 

The onslaught can lead to technology hype fatigue, especially in an era where 70% of digital transformation initiatives struggle. Most promise quick wins, but end up distracting employees from work, draining resources with long learning curves, and falling behind ambitiously set schedules. 

But one aspect of artificial intelligence is truly turning heads of leading insurers: Generative AI (Gen AI). In fact, 43% of insurance companies are using AI today—and for many, Gen AI is topping the list. 

What Is Generative AI in the Insurance Industry?

ChatGPT, Midjourney, and DALL-E are common newsmakers, but Generative AI is an all-around workhorse for insurers. While many new technologies are point solutions, promising improvements in a single area, Gen AI is a larger technology that can extend into all aspects of insurance operations. 

It’s more than a fun tool for penning wedding vows or recommending dinner dishes. It’s the ultimate ‘low-code’ tool, a moniker for business technology that doesn’t require a data science or programming degree to wield. Low-code tools are making tech easier for every employee to engage with. 

Generative AI is helping insurers:

• Analyze the vast amounts of data (written reports, transcripts, videos, and images) involved in the claims process
• Automatically generate new policies, quotes, and certificates
• Improve upsells by creating ad hoc promotions that appeal to hyper-specific user needs
• Detect fraudulent behaviors in submitted claims

Let’s see how Gen AI is improving all areas of insurance operations. 

Five benefits of Generative AI for insurers

Gen AI isn’t a specialist in one area—it infuses benefits organization wide. It’s a powerful technology that 90% of sales and marketing leaders believe their organization should use at least “often,” but businesses leaning into the technology are seeing the most gains. See how insurers are getting the most out of Gen AI. 

1. Improve Customer-Facing Chatbots

You’ll also find Gen AI powering conversational interfaces in the insurance industry. Chatbots and virtual agents are leveraging the technology to improve their communication skills. Whereas chatbots used to ace resetting passwords or looking up policy numbers, they’re now expert conversationalists. Customers can ask more nuanced questions of their policy, while insurers can boost productivity by 40–60% across their support desks.

The summarization power of artificial intelligence doesn’t just benefit employees, but your customers as well. Clear, accurate self-service answers are key to winning today’s customers: 22% would rather spend the night in jail than phone up a support line. Using Generative AI, insurance documents are no longer ‘static,’ but fully interactive. Customers can chat with their coverage to better understand benefits. Instead of toiling through a complex paragraph of jargon, customers can ask questions directly and receive accurate, ‘snack-sized’ answers. 

2. Reduce Manual Workload

Large-scale Gen AI adoption is boosting productivity by up to 30% for claim-related administrative tasks. Insurers are using it to automate mundane, time-consuming tasks like manual data entry and organizing incoming materials. 

3. Introduce a New ‘Search’ Paradigm

Ten years ago, McKinsey found that the average worker spent 1.8 hours per day searching for information. Since then, the amount of data churning through the average organization has doubled every two years. Tracking down internal information—from specific paperwork to supporting documents—can drain productivity exponentially. Generative AI will dramatically change how employees in data-laden industries search for and find information. 

Large Language Models (LLMs) make it easier for computers to ingest and interpret text. And not just what it says—but what it means. To find a file on a conventional intranet or internal data library, you’d have to use very precise keywords. But using a Gen AI-powered search engine, you can unearth what you need with more flexible, less precise, conversational language. 

For example, Gen AI understands the intent of your search. Instead of retrieving what you exclusively ask for, it can also serve other documents related to your query. Find out how many claims were approved in a specific region, what types of policies are the most popular amongst top salespeople, or find documents from similar claims to compare against. Information gathering becomes less of a hunt for a precise file, but a source of holistic answers. 

4. Generate New Documents from Existing Knowledge

Insurers are using Gen AI to automatically produce novel documents such as policy papers, insurance agreements, customer letters, and claim forms. Most remarkably, it can personalize policy documents for individual clients, tailoring them to their unique needs and histories.

5. Make Sense of Unstructured Content

Tech like Robotic Process Automation (RPA) has made it easier to manage and utilize structured data. Software can scan documents for information that follows a rigid format, like a phone number or policy ID. However, unstructured documents, those that follow looser formats, are notorious bottlenecks for both human workers and computers. Think images, videos, and written notes: content that’s vital to insurance operations but difficult to parse. 

Gen AI is solving the unstructured document problem for insurers—a boon for today’s organizations where 80–90% of data is unstructured. While data used to be locked inside images and video, Gen AI brings them to life. Insurers can ‘search’ the contents of multimedia and ask questions. When computers better understand more complex file types, they can also help us keep them better organized. 

Does Generative AI Carry Risks for Insurers?

Highly regulated industries have more to consider than retailers or hobbyists plunging into the world of Generative AI. Here are three industry concerns and how you can prepare to mitigate their risks to get the most out of Gen AI.

1. Privacy preservation: Insurance claims contain personal details that require close guarding. Gen AI can ingest a groundswell of information without revealing its source. That means you can leverage some of AI’s most useful tools with anonymized data. Simulate insights without revealing personal customer details, or model test scenarios to see how they’d impact coverage. Make sure vendors use AI, work with your data in a way that’s transparent, safe, and designed to protect data privacy (like Skan!) 
   
   
2. Humans-in-the-Loop (HITL) automation: While AI can accelerate decision making, it’s important to keep an eye on things. HITL views human staffers as an equal part of the automation picture, responsible for overseeing AI’s recommendations. AI models can save time—but left unfettered, they can sometimes learn unwanted patterns. Some cities, like New York City, have instituted new policies that require businesses to disclose how AI tools reach decisions. This reduces the likelihood of unwanted biases running rampant, like the employer that discovered it was accidentally discriminating against women applicants.
   
   
3. Data accuracy: Powerful AI tools like Skan can suggest the “next best action,” but only when they’re fed accurate data. When using Gen AI to fuel chatbots and internal knowledge bases, it’s important to keep your data lake fresh, updated, and complete with relevant business context. Gen AI can boost productivity and efficiency, but new roles will emerge; like those responsible for maintaining a clean feed of data. 

By planning ahead with these considerations in mind, you can turn this buzz-worthy technology into a winning strategy.

How to Get Started: Generative AI Use Cases in the Insurance Industry

Not sure what Gen AI strategy is best for your team? Here’s how two top insurers used Skan to help evaluate potential tools. 

Digital Twins Help De-Risk Generative AI Tools for This Insurance Company

Using Skan’s “digital twin” tech, one Fortune 100 insurer achieved over $10 million in savings. Amidst evolving global regulations, including the EU’s Artificial Intelligence Act, insurance companies recognized the need to test Gen AI tools for potential risk. 

Skan created a digital twin of one of its business operations—a real-time analytic representation of how their people work and their operations run. This proactive test lab helped this healthcare payor identify inefficiencies, make improvements, and improve throughput in the way its people and its AI agents route and process work. 

Evaluating the Effectiveness of Large Language Models in Underwriting Operations

Skan played a crucial role in helping an insurance carrier optimize its underwriting process by testing various Large Language Models (LLMs). They wanted to reduce reliance on traditional search engines, and create their own internal functionality fed with their own data. 

Skan installed real-time monitoring software to analyze the steps involved in the underwriting process. Participants were split into a series of groups, testing different LLMs for effectiveness. In the background, Skan technology closely monitored how users interacted with each product, and measured how each one impacted processing time. 

Ultimately, this Fortune 500 insurer had the data-driven insights they needed to pick the Gen AI product that worked best for their team—one that reduced processing time by 55%. 

Leading Insurers Are Winning with Gen AI: Are You One of Them?

Top carriers have automated more than half of claims activities. Gen AI isn’t just another newsmaker: it will reinvent how the insurance industry operates. Insurers at the forefront are moving away from the pack by reducing manual legwork, boosting productivity, and launching handy chatbots that customers love.

Request a demo to learn how your team can operationalize Gen AI.