How insurers can leverage the power of generative AI

The world of artificial intelligence continues to evolve rapidly, and generative AI in particular has sparked universal interest. This is certainly true for the insurance industry, where generative AI could fundamentally reshape everything from underwriting and risk assessment to claims processing and customer service.

The transformative power of this technology holds enormous potential for companies seeking to lead innovation in the insurance industry. Amid an ever-evolving competitive landscape, staying ahead of the curve is essential to meet customer expectations and navigate emerging challenges. As insurers weigh how to put this powerful new tool to its best use, their first step must be to establish a clear vision of what they hope to accomplish.

The role of generative AI in insurance

Generative AI, as its name suggests, generates content. Because its algorithms are designed to enable learning from data input, generative AI can produce original content - such as images, text and even music - that is sometimes indistinguishable from content created by people.

Models such as GPT 3.5 and GPT 4 present opportunities to radically improve insurance operations. They can automate processes, enhance customer experiences and streamline core functions across the insurance value chain. This ultimately will drive efficiency and effectiveness across the industry.

Use cases come into focus as more insurers adopt generative AI

As the number of companies exploring generative AI multiplies, so will industry use cases. Recently, four patterns have started to emerge among insurers.

1. Summarize policies, documents and other unstructured forms of content and data.
2. Synthesize summarizations to create new content and insights.
3. Answer questions based on what it has learned from summarizing and synthesizing.
4. Translate between natural languages, such as English and Italian as well as computer code; for example, translations that yield new modern code that can run on the cloud.

Three primary objectives

Insurers are seeking to leverage these expanded patterns to address three main objectives.

1. Improve experiences for customers, agents, agency staff members and employees by deploying a generative AI virtual assistant or virtual agent. Insurance companies can use generative AI to reinvent their approach to providing customer service and creating new experiences. Individualized and empathetic human interactions, for example, become easier to achieve when generative AI removes mundane processes from insurance professionals’ workload.
2. Heighten productivity and efficiency by deploying this technology alongside insurance industry knowledge workers, such as underwriters, actuaries, claims adjusters and IT engineers. For example, efficiency benefits include summarizing and synthesizing large volumes of content gathered during the claims lifecycle, including call transcripts, notes, and legal and medical paperwork. This is particularly useful in property and casualty insurance. Companies can compress the claims lifecycle dramatically. In the life insurance industry, there is significant interest in using generative AI for automation and decision-making in underwriting processes and policy issuance to a broader range of customers without the need for, say, in-person medical exams.
3. Manage compliance and mitigate risks, which are key concerns in this highly regulated industry. Automated compliance monitoring, fraud detection and even generating content in the form of training materials and interactive modules for staff to stay current on the latest regulations are areas that companies are starting to explore.

Risks and human oversight

While generative AI is valuable for identifying risks that humans overlook, the technology itself carries associated risks. These involve elements such as regulatory exposure, intellectual property, corporate-level reputation and bias, and information security. To mitigate such risks, insurers must embrace accountability and have control procedures and compliance frameworks in place. To ensure ethical and nondiscriminatory generative AI models, responsible AI methods that include human oversight are essential.

Three steps to help insurers get started

To drive better business outcomes, insurers must effectively integrate generative AI into their existing technology infrastructure and processes. Generative AI is a tool within a broader set of techniques and AI and automation technologies. Accordingly, insurers should improve existing processes and optimize them in parallel to achieve the maximum benefits of generative AI. The big win often involves combining multiple AI technologies to address different aspects of a project, such as semantic searching or language capabilities.

For insurers who are not yet using generative AI, these three initial steps are recommended.

1. Establish a multidisciplinary team of businesspeople, IT specialists and data scientists to embark on the generative AI journey, focused on adapting generative AI solutions to the unique requirements of the organization.
2. Identify the operating model that best fits the organization, not only in terms of experimenting with the technology but also to deploy it safely, successfully and in a scaled fashion.
3. Develop the requisite expertise and capabilities by beginning with low barrier use cases and gradually fine-tuning models based on domain knowledge and data sources.

With generative AI’s potential exceedingly clear, it is easy to understand why companies across virtually every industry are turning to it. As insurers begin to adopt this technology, they must do so with a focus on manageable use cases.

Chris Raimondo is Americas Insurance Technology Leader with EY. He may be contacted at [email protected].

Any opinions expressed here are the author’s own. These views do not necessarily reflect the views of Ernst & Young or other members of the global EY organization.

© Entire contents copyright 2023 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.