While most companies prefer to make final decisions on their own, Bestow lets the AI determine eligibility for a much faster process.
When most of us look at life insurance options, we touch base with our insurance agents and research other vendors. The process that ensues at the insurance company may involve a medical exam, many manual references of data, and a decision process that is performed by an underwriting expert.
SEE: TechRepublic Premium editorial calendar: IT policies, checklists, toolkits, and research for download (TechRepublic Premium)
The process can take several weeks and be costly to administer. Often, applicants may need to take a physical exam, provide detailed medical records, undergo a phone screen, and answer as many as 50 questions on a questionnaire.
“It shouldn’t take several weeks or months to buy anything, including your life insurance policy,” said Ben Hsieh, director of product development at Bestow, which offers online term life insurance. Hsieh said that in a little as five minutes, a life insurance applicant can apply online and, if approved, get coverage instantly without ever requiring a doctor’s visit, phone screening, or medical exam.
The process is driven by artificial intelligence (AI) that collects and operates on applicant-provided information and third-party data, including prescription history, motor vehicle records, insurance fraud and criminal history.
SEE: AI and machine learning are making insurance more predictable (TechRepublic)
“The type of data elements used by our AI expert systems isn’t particularly unique—it’s standard across the industry,” Hsieh said. “What’s unique is that we’re able to review approximately 3,000 data elements in real time to determine eligibility and pricing. By harnessing algorithms instead of manual underwriting approaches, we’re able to offer customers a fully in-force term life insurance policy entirely online in minutes.”
The AI process seems to be working. Hsieh reported that customers give his company a 4.8- out of 5-star rating on Trustpilot, a customer review platform.
“Our goal was to create an industry-leading AI underwriting platform that enabled us to expertly determine risk while also operating in a fully digital capacity,” Hsieh said. “In developing our proprietary underwriting technology, we’ve been able to identify which data elements and resources are truly needed to measure risk. And we supplement time-consuming touch-points with algorithms and data calls to expand eligibility, provide the best possible pricing for each applicant, and deliver a near-seamless buying experience.”
Unlike many AI platforms, which leave the final decision to a human expert, Bestow’s platform allows big data and algorithms to fully make the underwriting decisions for life insurance.
“In leaving the decision-making to the AI and its algorithms, we believe that we have solved one of the major challenges when human underwriters make these decisions,” Hsieh said. “In human decision-making, there can be inconsistencies and bias.Two human underwriters could look at a customer profile and come to two separate decisions based on their judgment and experience. An AI-powered underwriting approach is more consistent, more accurate, and you can generate valuable structured data to improve the underwriting over time.”
That is not to say that AI decisions are without bias. What AI can do is to render the data that gets into decision-making cleaner and more accessible.
SEE: Natural language processing: A cheat sheet (TechRepublic)
To combat bias in the AI, data models must be continuously reviewed and revised.
“AI cannot, at least at this point, replace the qualitative analysis of a person,” Hsieh acknowledged. “We also recognize that by relying on a 100% digital underwriting platform for our coverage decisions, it can admittedly be challenging to create decision-making logic for complex applicant cases, where usually a human underwriter would be flagged to review individual risks. In addition, there are limitations in available real-time data replacing traditional information available from medical records and exams. While we don’t think this will be a long-term limitation, it has forced us to be creative in using data and defining rules to provide as many people as possible with access to coverage.”
AI decision making is a work in progress for Bestow and many other companies that will improve over time. But meanwhile, consumers’ own times-to-decision are being shortened by automated AI processes. There’s value in that.