Manulife

Canadian life insurance applicants historically faced a cumbersome underwriting process characterized by lengthy medical questionnaires covering broad "body system" categories — cardiovascular, respiratory, and similar — regardless of individual risk profiles. Every application, regardless of complexity, passed through manual underwriter review, creating bottlenecks that delayed coverage access by days or weeks. For advisors guiding clients through these forms, the friction eroded the client experience and increased administrative burden. In an industry where speed of protection matters — families left uninsured during review periods face real financial exposure — the status quo imposed both operational costs and customer satisfaction penalties at scale across Manulife's Canadian distribution network.

Manulife enhanced MAUDE (Manulife Automated Underwriting Decision Engine), its proprietary predictive ML underwriting engine first introduced in 2018 as AIDA — the first AI tool in Canada to make automatic underwriting decisions. The 2025 enhancement paired MAUDE with a fully redesigned electronic application built around reflexive, adaptive questioning: the form dynamically adjusts which medical questions are presented based on the applicant's age, coverage amount, and prior responses, eliminating irrelevant body-system question sets. Standardized medication and condition pick lists, drop-down menus for travel and hobbies, and paramedical exam flexibility were integrated to improve data consistency and processing speed. Applications that cannot be automatically approved are seamlessly routed to human underwriters, preserving personalized consideration for complex cases. The rollout launched in fall 2025 across Manulife's Canadian advisor network, with training and demos made available through the Manulife Advisor Portal.

By December 2025, MAUDE was automatically approving 58% of eligible cases — a 56% increase over pre-launch levels — with qualified applicants receiving decisions in as little as two minutes. The redesigned application reduced medical questions by up to 40%, cutting completion time for advisors and applicants alike. Advisor adoption was described as strong shortly after launch, with the majority of eligible submissions flowing through the automated path within weeks of rollout. Qualitatively, the reflexive question format reduced back-and-forth clarification, and standardized pick lists improved data consistency feeding into MAUDE's decisioning model. Manulife's broader AI program, active since 2016, is expected to generate over $1 billion in enterprise value by 2027.

• Adaptive, reflexive questioning upstream of the AI model directly improves model input quality — cleaner, more consistent data produces faster and more accurate automatic decisions.
• A defined escalation path to human underwriters is not a fallback; it is a design requirement that maintains regulatory trust and handles edge cases without degrading the applicant experience.
• Proprietary AI built on years of production data compounds in value — MAUDE's 2025 gains reflect seven years of iterative learning unavailable to insurers relying on off-the-shelf models.
• Advisor-facing tooling (training resources, portal demos) is a critical adoption lever; straight-through processing rates depend on advisors submitting applications correctly through the new workflow.
• Responsible AI governance frameworks should be embedded from the start, not retrofitted — Manulife's published AI principles provided the foundation for expanding automation without sacrificing auditability.