How Will Two New AI Chiefs Transform Zurich Insurance?
Zurich Insurance Group has named Terry Powell as Group Chief Technology Officer and Cristina Ghetti as Group Head of Digital Employee Experience. Both start on 20 October 2025 and report to Ericson Chan, Group Chief Information & Digital Officer.
The signal is clear: import customer-first AI leadership from retail and fintech to speed up execution. Zurich is moving from pilots to scaled delivery across its core operations.
Why these hires matter
Terry Powell previously led IT transformation at ANZ Bank and AI-led customer initiatives at Domino's Pizza. His mandate: embed AI in core infrastructure to increase resilience, performance, and efficiency.
Cristina Ghetti brings large-scale tech adoption experience from Nestlé, where she helped oversee one of the sector's biggest AI rollouts. Her brief: build AI fluency across Zurich's workforce so teams can apply tools responsibly and at speed.
As Ericson Chan notes, the goal is to accelerate digital transformation with a customer-centric approach that improves experiences for both customers and employees.
From experimentation to enterprise integration
Zurich has tested widely. Its Agentic AI Hyper Challenge produced more than 200 prototypes aimed at real industry problems. The company has also partnered and invested across insurtech to widen its option set.
These appointments indicate a shift to industrialization: fewer experiments, more production-grade platforms, clearer governance, and measurable outcomes.
Where AI will move the needle in insurance
- Claims: Triage, severity prediction, document extraction, and straight-through settlement for low-complexity cases.
- Fraud: Network analysis, anomaly detection, and LLM-supported investigation to cut false positives and losses.
- Underwriting: Risk signals from unstructured data, better pricing segmentation, and faster quote-to-bind.
- Billing and operations: Automation of reconciliations, collections, and service workflows.
- Customer and agent service: AI assistants for inquiries, policy changes, and knowledge retrieval.
- Employee enablement: Training, prompts, and safe toolkits that fit into daily workflows.
Large language models can analyze vast datasets to improve risk assessment and flag fraud early. The payoff: shorter cycle times, better loss ratios, and lower expense ratios-if governance keeps pace.
Managing new risks
AI is now part of how insurers prepare for emerging exposure, including autonomous vehicles and cyber-physical risks (for example, vehicle system hacking). That calls for updated liability models, new data sources, and scenario testing.
Strong guardrails reduce model risk and bias. For reference, see the NAIC AI Principles and the NIST AI Risk Management Framework.
What to watch at Zurich
- Claims: Cycle time, straight-through processing rate, and loss adjustment expense.
- Fraud: Detection lift, false positive rate, and recovery value.
- Underwriting: Quote turnaround, hit ratio, and pricing accuracy.
- Operations: Cost per policy, automation coverage, and service levels.
- Workforce: AI adoption rates, productivity gains, and compliance audit results.
- Platform health: Model drift, latency, uptime, and incident frequency.
Action checklist for insurance leaders
- Prioritize 5-7 use cases with clear P&L impact and a 6-12 month payback path.
- Stand up data contracts for the sources those use cases need (quality, lineage, consent).
- Define guardrails: model validation, human-in-the-loop, privacy, security, and audit trails.
- Equip teams: role-based training, prompt standards, and approved toolkits inside core systems.
- Move from pilots to platforms: shared services for LLMs, MLOps, monitoring, and access controls.
- Set leading indicators early and report weekly-fix or retire what underperforms.
- Use a partner mix: cloud, model providers, and insurtechs-measured against business KPIs.
- Run change management like a product: clear narratives, quick wins, and steady enablement.
Upskilling your team
If you're building AI fluency across underwriting, claims, and operations, consider structured learning paths. Explore curated options here: AI courses by job.
Bottom line
Bringing in leaders from retail and consumer sectors suggests Zurich wants speed, simplicity, and measurable outcomes. If execution follows, expect tighter operations, better customer experiences, and a workforce that knows how to apply AI where it counts.
Your membership also unlocks:
