As a cornerstone of the insurance industry’s digital evolution, Simon Glairy has spent years navigating the intersection of risk management and emerging technologies. His work focuses on transforming traditional corporate structures into agile, AI-driven powerhouses that prioritize both human talent and technical precision. In this conversation, we explore how large-scale organizations can move beyond theoretical innovation to embed intelligence into the very fabric of their global operations. We discuss the mechanics of large-scale workforce upskilling, the shift toward agentic AI in claims processing, and the rigorous educational frameworks required to maintain a competitive edge in a rapidly shifting technological landscape.
When rolling out a digital initiative to over 144,000 employees across 70 countries, how do you balance global standards with local operational needs? Please describe the specific milestones that indicate AI tools are being embedded into daily workflows rather than remaining theoretical concepts.
The challenge lies in creating a unified vision that still breathes within local contexts, ensuring that a claims adjuster in one region feels the same empowerment as a data scientist in another. We achieve this through a mix of expert-led masterclasses and role-specific education that transforms AI from a distant concept into a practical daily companion. A critical milestone is when we see teams move from general curiosity to developing concrete, localized solutions that address their specific pain points. The true indicator of success is the “Global AI Run” philosophy, where the tool is no longer an “extra” task but an integrated part of the business agility framework. By involving 144,000 members of the team, we ensure that the rollout isn’t just a top-down directive, but a bottom-up adoption that prioritizes inclusivity and tangible impact.
Using “value creation” as the primary metric for innovation requires a rigorous framework. How do you quantify the long-term return on internal training programs compared to direct software implementation, and what steps ensure that efficiency gains remain sustainable?
Value creation is our North Star, but it must be measured through the lens of human capability as much as software performance. While direct software implementation offers immediate features, our internal training programs, such as the Data Excellence and Fit4IT initiatives, build the foundational “buy-in” that prevents technology from becoming shelfware. We look at sustainability by monitoring how internal culture shifts toward self-sufficiency, reducing the need for constant external intervention. When employees understand the “why” behind the tool, they become the architects of their own efficiency, ensuring that gains are not just one-time spikes but permanent shifts in operational excellence. This careful, measured approach ensures that every dollar spent on a person provides a compounding return that a software license alone simply cannot match.
Agentic AI frameworks manage low-complexity tasks while specialized ecosystems handle document synthesis. How do these tools change the specific skill sets required for entry-level roles, and can you share an anecdote of how an employee’s daily routine shifts once these automations are fully active?
The introduction of frameworks like Project Nemo has fundamentally redefined the entry-level experience by stripping away the “drudge work” of low-complexity claims. Previously, an employee might spend hours on manual data entry and basic file sorting, but now, with an 80% decrease in the time required to process and settle these files, their day is transformed. Imagine a claims handler who used to be buried under paperwork; today, they use AllianzGPT to synthesize complex documents in seconds, allowing them to focus on high-touch customer empathy and complex problem-solving. This shifts the required skill set from rote processing to advanced prompting, data synthesis, and critical decision-making. The daily routine moves from being a “cog in the machine” to being an orchestrator of AI tools, where the human touch is reserved for the moments that truly matter.
Establishing nine-month industrial learning programs for technical specialists involves a significant time commitment. Why is this deep academic level of training necessary for data engineers today, and how do you ensure the curriculum keeps pace with technology that evolves faster than traditional university cycles?
In a field that moves as fast as AI, a weekend workshop is a band-aid, not a solution; that is why we partnered with Sorbonne University for a formal nine-month industrial learning program. This depth is necessary because data engineers must understand the underlying architecture of machine learning, not just how to use a specific interface. By embedding our specialists in an academic environment, we ensure they are operating at the cutting edge of theoretical science while applying it to real-world insurance data. To keep pace with rapid shifts, the curriculum is designed to be a “living” syllabus that blends foundational principles with the latest industry breakthroughs. This ensures our engineers aren’t just reacting to the next tech wave, but are equipped with the analytical rigor to drive it themselves.
Executive-level training often moves beyond general literacy into long-term business modeling. What are the most common friction points when challenging senior leaders to integrate AI into their strategy, and how do you move them from basic awareness to active, impact-driven deployment?
The most common friction point for senior leaders is often the “black box” syndrome—the hesitation to rely on a technology they don’t fully see the mechanics of. We move them past this by shifting the conversation from “what AI is” to “how AI redefines our business model” through specialized executive tracks. This transition requires showing them that AI is not a tech project, but a strategic lever that affects everything from procurement to organization. We provide them with the tools to build long-term business models where AI is a core pillar of growth, rather than an experimental add-on. By focusing on active, impact-driven deployment, we turn skepticism into leadership, where executives become the primary advocates for a future-ready, digital-first culture.
A tiered learning matrix allows employees to upskill at their own pace, from ethics to advanced machine learning. How do you maintain high engagement across such a diverse workforce, and what metrics do you use to track the transition from foundational literacy to expert-level application?
Engagement is maintained by providing a clear, structured roadmap that respects the different starting points of 144,000 individuals across 70 countries. Our learning matrix starts with basic AI fundamentals and ethics—ensuring everyone speaks the same language—before progressing into advanced prompting and specialized machine learning. We track this transition through the “Global AI Run” milestones, observing how many employees move from foundational literacy modules to developing their own concrete solutions. We also measure the acceleration of AI tool adoption across different departments as a proxy for engagement and expertise. When we see a “perfect symbiosis” between our technology teams and our business units, we know the tiered system is working and that the skills are being applied in real-time.
What is your forecast for AI in the insurance industry?
I believe we are moving toward a “zero-latency” insurance model where the gap between a claim being filed and settled becomes nearly invisible for low-complexity cases. In the coming years, agentic AI will not just assist humans but will proactively manage entire claim lifecycles, allowing our workforce to transition into roles that are entirely focused on high-level risk strategy and complex human advocacy. We will see a shift where “AI literacy” is no longer a specialized skill but a basic requirement for every professional, much like using a computer is today. Ultimately, the companies that thrive will be those that view AI as a tool for human empowerment, using efficiency gains to reinvest in deeper, more meaningful customer relationships. Success will be defined by how well we marry the speed of the machine with the trust and integrity of the human expert.
