Insurers rely heavily on data for risk assessment, pricing optimization, and claims management, but around 80% of all insurance data lies trapped in everyday correspondence, including submissions, legal demands, and medical records.
The information within this unstructured data is frequently incompletely analyzed, misread, or unused, costing insurance companies tens of billions of dollars annually due to underpricing premiums, insurance fraud, and overpaying claims. The insurance industry’s reliance on human experts to carry out non-core and administrative activities, such as manually entering and reworking information from documents, will contribute to an estimated $85 billion to $160 billion in value lost to inefficiency by 2027, according to Accenture.
Generative AI provides an opportunity to avert these losses by maximizing efficiency and effectiveness across underwriting and claims operations. Accenture forecasts that generative AI will automate up to 62% of underwriting and claims processes. Reducing the manual effort associated with accessing and assessing this data is a critical step toward increasing the value created by human experts. It allows experts to concentrate exclusively on higher-value tasks that drive good decision-making. Additionally, according to Swiss Re, by using AI to extract insights from unstructured data sources, insurers can see a 12% to 25% improvement in their loss ratio compared with companies that don’t use this technology.
What are other potential benefits from generative AI in underwriting and claims?
See also: Cautionary Tales on AI
AI Impact on Underwriting
An insurance-specific AI solution provides several benefits. The first is faster speed-to-quote by accurately extracting data from submissions with 95%-plus accuracy while keeping up with broker demand to bolster the customer experience for policyholders. AI also increases capacity and lowers administrative overhead, reducing the time underwriters spend manually reviewing and inputting data from unstructured sources.
Generative AI designed for insurance drives better profitability by surfacing more information within unstructured documents, which can reduce error rates and premium leakage by 1% to 3%. Lastly, the technology can improve talent retention to mitigate the effects of insurance’s talent crisis by reducing agents’ tedious, administrative work and empowering them to focus on more strategic, engaging activities.
Claims Benefits From AI
Insurance-specific AI lets claims teams use insights from unstructured data to reduce claims costs across the entire process, from first notice of loss through claims recovery opportunities. It also enhances customer experiences by accelerating claims document handling time by 90% to settle claims faster.
The technology lets insurers increase their capacity to contend with spikes in claims submissions driven by growing natural catastrophe risks. It enhances security and reduces compliance overhead through greater quantities of decision data, which can yield insights to help insurers adhere to full process transparency. Thus, AI enables greater cost control while lowering users’ risk exposure.
See also: Who's Getting Results From AI, and Why?
How can insurance companies get started with AI? Below are some steps for success:
- Select a pilot project by launching a small-scale implementation in a specific use case for a single product or line of business. This allows for testing and refinement before a broader rollout—or to “fail fast” if the value proposition is unclear.
- Prepare your data to ensure the accessibility of unstructured data. This may involve digitizing paper documents and centralizing data from various sources.
- Plan your integrations by closely coordinating with IT teams to plan how AI will integrate with existing systems and workflows. Review APIs and other integration strategies.
- Employ training and change management to prepare teams for AI tools. This includes training on how to use the new systems and understanding “the change curve” to manage a cultural shift toward AI-augmented work.
- Partner with legal and governance/compliance teams to ensure that AI use aligns with regulatory requirements, internal risk management policies, and broader company values.
- Enable continuous monitoring and improvement by setting up systems to monitor the performance of AI-augmented processes, gather user feedback, and continuously refine the model based on new data and insights.
The insurance industry faces a significant challenge with unstructured data. However, the opportunity is equally massive for businesses that effectively harness this information. AI uniquely developed for insurance companies’ specific needs empowers insurers to easily and effectively transform unstructured data into structured data, informing underwriting and claims decisions and driving efficiency for maximum business impact.
