Manulife Financial

Life insurance underwriting has historically been one of the most friction-heavy processes in financial services. Manulife's individual insurance application required clients and advisors to work through a lengthy, undifferentiated questionnaire covering a broad range of health impairments — most of which were irrelevant to any given applicant. The result was a process that frustrated clients accustomed to digital-first experiences, burdened advisors with administrative work, and slowed approvals. In an industry where the pandemic had already demonstrated that many applications could be assessed without medical exams or fluid testing, the gap between client expectations and operational reality had become a competitive liability.

Manulife redesigned its electronic life insurance application and enhanced MAUDE (its AI underwriting tool, in production since 2018) to address both the data collection and risk assessment problems simultaneously. The application was shortened by removing approximately 40% of questions, replacing a generic impairment checklist with personalized, condition-specific questions focused on the four most prevalent health impairments in the applicant pool: heart disease, cancer, stroke, and diabetes. Follow-up questions drill into medication and treatment specifics to surface related conditions. The enhanced MAUDE system — used by Manulife underwriters, not advisors — applies predictive ML models to automatically underwrite younger, healthier applicants based on product type and coverage amount, routing only complex or ambiguous cases to human review. The redesigned application launched in fall 2025.

By December 2025, the combined application redesign and enhanced MAUDE system delivered measurable gains across speed and throughput:

• 58% of eligible applications were approved automatically through AI underwriting
• ~2 minutes approval time for healthy clients with no critical or chronic illnesses
• ~15 minutes faster per application for advisors compared to the prior process

Cases that fall outside the automatic approval criteria are routed to underwriters for manual review and can still result in standard approvals — the model flags uncertainty rather than rendering a denial. Advisors reclaim time previously spent navigating general impairment lists, redirecting it toward client advice and outreach.

• Pre-launch validation is essential: Manulife analyzed hundreds of applications side-by-side before rollout, confirming that removing 40% of questions did not create gaps in collected medical information.
• Targeted questioning outperforms exhaustive questionnaires — focusing on the four highest-prevalence impairments with medication-level follow-up captured equivalent clinical detail in less time.
• AI underwriting works best as a triage layer: routing straightforward cases to automation while preserving human review for edge cases maintains underwriting quality without sacrificing speed.
• Adoption metrics should be defined before launch; tracking the share of eligible applications auto-approved provides a clear signal of system effectiveness and flags any drift over time.