Swiss Re expands AI mental health initiative: what insurers should take from it
Swiss Re and Wysa have scaled their digital mental health program, Wysa Assure, to policyholders across three continents. The app blends an AI chatbot built on CBT principles with Swiss Re's risk expertise and proprietary risk scoring. The goal is straightforward: earlier support, less symptom escalation, and a path to lower claims exposure.
Launched in 2023, Wysa Assure gives users on-demand emotional support, self-guided resources, and clinically validated modules. For insurers, it offers a prevention-first layer that can plug into existing benefits, wellness, and claims strategies.
What's new
- AI-led support grounded in CBT with self-guided, clinically validated pathways.
- Integration with Swiss Re risk scoring to inform proactive outreach and engagement.
- Positioned for early intervention, aiming to reduce severity and duration of mental health episodes.
Where it's live
- Ascenda (Australia) - adoption led to Health & Wellness wins in Group and Individual Life at the Financial Standard Industry Service Award - Life Insurance 2025.
- MassMutual (United States)
- Eagle Insurance (Mauritius)
- Momentum (South Africa) via the Multiply Rewards Programme, making access available to all policyholders.
Swiss Re frames these partnerships as part of a wider shift: insurers are using proactive health tools to lift engagement and manage risk earlier in the journey. Done well, this can smooth utilization, reduce avoidable claims, and improve retention.
Why this matters for underwriting, claims, and distribution
- Underwriting and risk: Engagement and symptom trends can inform dynamic risk views while guarding against bias and selection effects.
- Claims: Earlier support may reduce incidence and severity for mental health, disability income, and absence claims, while improving return-to-work timelines.
- Distribution and retention: Mental health support packaged in rewards and wellness programs boosts perceived value, NPS, and persistence.
- Compliance and data: Clear consent, data minimization, and governance are essential across GDPR/POPIA/HIPAA contexts. Keep model risk management in scope.
The Momentum example
Momentum offers Wysa through Multiply, meeting a large access gap. Research indicates roughly one in three South Africans will experience a common mental disorder in their lifetime, underscoring the need for scalable support tools.
"Complete health calls for more than physical care - it's also dependent on sleep, human connection, and mental wellbeing. Our partnership with Wysa enables our members to manage their mental health in a new way with effective and accessible mental health resources, ensuring holistic care," said Maria Pouroullis Carpenter, head of Momentum Multiply.
Efforts like this also help reduce stigma by normalizing mental health as part of everyday benefits. For context on mental health and care approaches, see overviews from the World Health Organization and CBT guidance from the NHS.
How to evaluate an AI mental health partner
- Clinical safety: Evidence base, clinical oversight, and crisis escalation pathways to human support.
- Data safeguards: Encryption, access controls, audit trails, and third-party certifications.
- Localization: Language, cultural fit, and regulatory alignment across markets.
- Integration: APIs/SDKs for benefits platforms, rewards engines, and claims systems.
- Outcome proof: Peer-reviewed results or validated measures demonstrating symptom reduction and engagement.
- Commercial model: Pricing tied to activation and outcomes, not just licenses.
- Governance: Model documentation, bias testing, red-team processes, and explainability where feasible.
Pilot framework you can run this quarter
- Pick a segment: income protection, group life, or employer-sponsored benefits.
- Define success up front: enrollment, 30/90-day active use, symptom improvement, claim notices, and RTW timelines.
- Configure clinical guardrails: screening, escalation to human care, and crisis handling.
- Run a 90-120 day pilot with a matched control group where possible.
- Review results, data quality, and member feedback; proceed to staged scale-up if targets are met.
KPIs that matter
- Eligibility-to-enrollment rate and activation within 7 days.
- 30/90-day active use and completion of modules.
- Changes on validated measures (e.g., PHQ-9, GAD-7) when appropriate and consented.
- Time to first support interaction and repeat engagement.
- Claim notice rate and duration vs. control; sick-leave days for group products.
- Member NPS, complaints, and lapse/persistency shifts.
- Cost per engaged user and estimated medical/claims offsets.
Risks to manage
- Over-reliance on a chatbot without timely escalation to clinicians when needed.
- Misclassification or delayed triage in higher-risk cases.
- Equity concerns if engagement varies across demographics.
- Regulatory exposure without explicit consent and clear data boundaries.
- Brand risk if marketing overpromises clinical outcomes.
Bottom line
Mental health support is moving from a nice-to-have to a baseline feature in insurance. The Swiss Re-Wysa expansion signals where the market is going: build proactive access, integrate risk signals responsibly, measure outcomes, and scale what works. Start with a tight pilot, hold the vendor accountable to data, and connect engagement to real claims and retention impact.
If your team is building AI fluency for insurance use cases, explore role-based learning paths at Complete AI Training.
Your membership also unlocks:
