Canadian insurtech PolicyMe says its new in-house AI copilot has reduced underwriter turnaround time on complex life insurance applications by 59%, tackling a process that can leave applicants waiting while medical records, lab results and physician statements are reviewed.
While most PolicyMe applicants already receive decisions in seconds, cases involving more detailed medical histories require closer scrutiny. The copilot brings supporting documents into one workspace, organizes information for review and drafts plain-language explanations for customers. Underwriters retain responsibility for every final decision.
The reported improvement is based on PolicyMe’s internal data and measures the period from an underwriter receiving a complex case to completing the review, rather than the entire application process.
Built within PolicyMe’s own isolated cloud environment, the tool also allows underwriters to trace information back to its original source, with safeguards designed to prevent customer data from being used to train the underlying AI models.
Fintech.ca spoke with Jeff McKay, PolicyMe’s CTO and Co-Founder, about why the company focused on complex cases, how it protects sensitive health information, and where human judgment remains essential.
Before we dive in, how would you explain what the copilot does?
JM: Our new AI copilot is built for one of the more time-intensive parts of life insurance: reviewing complex applications. When an application involves a detailed medical history, underwriters often have to work through lengthy medical records, lab results and physician statements spread across several sources before they can make a decision. That work takes time, and while it happens, the applicant waits.
The copilot brings those records and supporting documents into one workspace, helps organize information for review, and drafts plain-language explanations for customers. Underwriters spend less time moving between files and piecing the case together by hand, so they can focus on the judgment calls that need their expertise. The underwriter still makes the final decision, and the copilot helps them work from a clearer, more complete picture.
Most applicants already receive decisions in seconds. What problem were you trying to solve with this tool?
JM: Most of our applicants already receive an underwriting decision in seconds. However, some applications aren’t as straightforward and require a closer look at medical records, lab results, and other information before an underwriter can make a decision. Reviewing that material can take days, particularly when there are multiple documents to work through, and the applicant waits in the meantime.
The copilot is designed to reduce the time spent organizing and reviewing those documents. That gives underwriters more time to focus on the substance of the case and the decision they need to make. We’re looking to improve the experience for customers with applications that need more attention, while keeping the fast process that already works for most applicants.
Why use a copilot for this work instead of fully automating the decision?
JM: Complex underwriting requires human judgment. The information in an application needs to be considered in context, and the underwriter is responsible for assessing the case and making the final decision. The copilot can help organize information, but it cannot take over that responsibility.
The role we’ve given the AI is to remove friction from the work. Underwriters can get through the information more efficiently, while the decision stays reviewable and accountable. We’re using the technology to support the people doing the underwriting, rather than to replace their judgment or decision-making.
PolicyMe says the copilot has significantly cut underwriting turnaround time. What does that measure?
JM: It measures the time from when an underwriter receives a complex case to when the review is completed. The result is based on PolicyMe’s internal data on complex cases, and reflects the time saved during the underwriter’s review.
How do you make sure underwriters can trust what the copilot produces?
JM: Underwriters review and approve everything the copilot produces before it reaches a customer. That review is an important part of the process. The copilot can organize the information, but the underwriter needs to be able to assess the output and take responsibility for what is shared.
We’ve also built traceability directly into the copilot. Underwriters can trace the information it presents back to the original medical records and supporting documents, allowing them to verify information at its source rather than relying solely on the AI’s interpretation. This helps make the copilot’s outputs more transparent and easier to validate.
Beyond that, we take a rigorous approach to testing. Before deploying new versions, we evaluate their performance against historical underwriting cases to assess accuracy and reliability, identify areas for improvement, and build confidence in the information the copilot provides.
What safeguards are in place for customer information?
JM: Protecting customer information is a fundamental part of how we’ve designed the copilot. It operates within our secure, isolated cloud environment, where sensitive information is handled within established security boundaries. We’ve also implemented technical safeguards to prevent customer data from being used to train the underlying AI models.
This is especially important in life insurance underwriting, where applications often contain highly sensitive medical and personal information. Our approach is designed to give underwriters the benefits of AI while maintaining strong controls over how customer information is accessed, processed, and protected.
What could this change for someone applying for life insurance?
JM: If an application needs a closer look, a faster review can mean the applicant hears back sooner. Long waits are a common frustration in complex cases, so reducing time spent in review is an important step in easing the wait for customers.
The copilot can also draft plain-language explanations for declined, postponed or rated applications. These outcomes can be hard to understand, so clear explanations are an important part of the experience for applicants. The underwriter is still responsible for the decision and for the explanation the customer receives.
Leave a Reply