Zurich Insurance launches AI Lab to solve real problems in insurance
Date: October 30, 2025
Zurich Insurance Group has launched the Zurich AI Lab, a focused push to build scalable AI solutions for practical challenges across the insurance value chain. The initiative is led by Group CEO Mario Greco and anchored in three locations: Singapore, Zurich and St. Gallen.
The lab pairs Zurich's industry experience with academic research through partnerships with the University of St. Gallen's Institute of Management & Strategy and ETH Zurich's Agentic Systems Lab. A multidisciplinary team will operate under the guidance of senior Zurich executives and faculty from both universities.
Leadership and partners
Executive sponsors include Group Chief Transformation Officer Carlos Rey de Vicente and Group Chief Information and Digital Officer Ericson Chan. Academic partners include University of St. Gallen professor Karolin Frankenberger and ETH Zurich professor Elgar Fleisch.
Fleisch noted that ETH students are already turning ideas into real AI applications, and the lab bridges that talent with a leading insurer to build what's next. Frankenberger highlighted the shared work on business model innovation and incumbent disruption, aiming to deliver research that moves the industry and society forward.
Greco said AI is already helping Zurich serve customers faster and provide more accurate risk insights. He described the AI Lab as a "moonshot factory" combining business expertise, technology, and research to reinvent how Zurich operates and set up the next generation of solutions.
What this means for insurance professionals
- Underwriting and pricing: Scalable models can pre-fill data, score risk at new levels of granularity, and improve portfolio steering. Expect faster quote cycles and tighter loss-cost projections.
- Claims: Smarter triage, quicker FNOL processing, document summarization, and better fraud flags. The goal is cycle-time reduction without losing control of leakage.
- Risk engineering: Consistent insights from unstructured data (photos, PDFs), pattern detection across accounts, and clearer guidance for insureds.
- Distribution and service: AI copilots for brokers and agents, cleaner handoffs across channels, and more accurate, compliant responses to complex queries.
- Operations and controls: Model monitoring, audit trails, and enforceable policies will be part of the lab's work-key for regulators and internal risk teams.
How to plug in-practical next steps
- Pick 2-3 high-friction use cases: e.g., submission intake, loss runs extraction, claims triage. Keep scope narrow and measurable.
- Get your data house in order: Define golden sources, access controls, and data lineage. You'll move faster later.
- Stand up model risk management: Set policies for validation, drift monitoring, bias testing, and human-in-the-loop review.
- Build small cross-functional squads: One business owner, one data scientist/ML engineer, one software engineer, one compliance partner.
- Measure what matters: Time-to-quote, bind rate, claim cycle time, severity, leakage, and NPS. Tie wins to P&L.
- Augment, don't over-automate: Keep expert judgment where it counts. Let AI handle repetitive work.
- Privacy and security by design: Sensitive data stays protected. Log everything. Make reviews repeatable.
Research outputs to watch
The Zurich AI Lab plans to publish studies on how AI affects insurance, strategy, and business models-grounded in data from real use cases. Expect applied insights you can reference with boards, regulators, and partners.
Context: momentum continues
In August 2025, Zurich North America introduced a new framework to support large-scale construction projects. Combined with today's announcement, the signal is clear: AI-backed operating models and targeted frameworks are moving from pilots to scalable programs.
Where to learn more
- ETH Zurich background and research updates: ethz.ch
- Upskilling options by job function (useful for underwriting, claims, and IT teams): Complete AI Training - Courses by Job
Bottom line
Zurich's AI Lab is built for scale and measurable outcomes. If you lead underwriting, claims, distribution, or operations, now is the time to pick focused use cases, set guardrails, and prove ROI-then expand with confidence.
Your membership also unlocks:
