AI agents may improve underwriting process, close talent gaps
Underwriters train AI agents to think like them, teaching the software to perform menial tasks, then larger projects.
Many underwriters have large workloads that require long hours and limited access to the data they need to make a decision quickly. Autonomous AI agents are here to help ease the burden on underwriters by performing menial tasks the same way an underwriter would so they can focus on more complex assessments.
AI agents act as assistants for underwriters and other insurance professionals and are intelligent entities that learn to make decisions and execute actions without direct human involvement. The technology can complete various tasks that could revolutionize the underwriting workflow, from generating accurate responses for risk quality questions to analyzing a customer’s claims history and business information. These AI agents can monitor a client’s activities and ultimately make a calculated recommendation of whether to accept or decline an applicant.
“The time it takes from a request to come in, to a quote getting insured is massively reduced [with autonomous AI agents],” said Sathish Kumar Manimuthu, chief technology officer at NeuralMetrics, creators of an AI-powered data platform that offers actionable risk assessments for P&C carriers, brokers and agents.
AI agent benefits
Some aspects of underwriting are tedious and take time away from other job duties that require more research and thought. AI agents can sift through data quickly, finding client information instantly, even for small businesses with limited digital footprints that usually require long, manual searches. The software saves underwriters time by finding incorrect information and inconsistencies quickly and efficiently and filtering the information for risk assessments and classifications.
“The intuition of an underwriter that’s developed over a period of time is not something that’s easily replaced,” said Manimuthu. “The AI agent framework is something that can replicate an underwriter persona… the underwriter can train the agent to start drawing inferences like how he would draw inferences.”
Changing and adding new risk metrics has been challenging for underwriters, but autonomous AI agents can more easily add and capture new risk characteristics, says Manimuthu. He gives the example of evaluating businesses open past midnight but wanting to tighten that data to assess risk for only businesses open after 2 a.m. “It’s very easy to teach the [AI] agent to look for those certain aspects, and the agent learns the capability.”
Filling talent gaps
Many industries are facing talent gaps, including underwriting and other insurance positions. A white paper from Attensi reports 55% of insurance executives surveyed reported talent acquisition and retention difficulties that could impact the company’s growth in the next year. This problem may only intensify as aging insurance professionals transition to retirement in the coming years. Leveraging AI technology may help overcome the talent crisis and attract high-impact employees to the P&C sector who embrace the industry’s evolving landscape and AI’s part in it.
“The most menial task that an underwriter is supposed to do, like portfolio management where you’re auditing certain parts of your portfolio… you can train an [AI] agent to do that specific activity, and they can rip through your entire book because they don’t sleep,” said Manimuthu. He adds that freeing up underwriters from these more monotonous steps allows them to focus on more complex activities and books that require more attention. “[AI agents] rapidly build capacity to the organization so they can underwrite much faster and more efficiently.”
Incorporating AI
Underwriters utilize many tools and computer software technologies to analyze risk and make a calculated decision. Autonomous AI agents are another tool to help underwriters streamline their workflows and move through risk assessments faster and more efficiently. However, incorporating AI into the everyday underwriter’s workflow will take time, and it doesn’t stop there. This type of software can streamline many areas of insurance including audits.
“These autonomous AI agents are literally nothing but pieces of software that can be automatically taught certain skills,” said Manimuthu. “They’re not like a silver bullet [where] you deploy it, and everything starts working from day one… it will evolve over a period of time as it learns more and more skills.”
Related: