The Digital Insurer founder on why the gap between pilot programs and enterprise deployment comes down to the ‘frozen middle’ — and what separates awareness from actual adoption.

Hugh Terry has spent over 12 years helping insurers around the world understand what digital transformation requires from a practical standpoint, not a theoretical one.

In this time, he has observed a consistent and frustrating pattern. Typically, AI adoption starts with executive enthusiasm for successful proof-of-concept projects, and genuine technological capability. Then this honeymoon period stalls, somewhere between the pilot and production.

Notably, it is not at the technological level that outcomes stall. Rather, there’s a space Terry calls the “frozen middle” — that layer where middle managers control how work actually gets done day-to-day. Their teams are measured on current performance, and for them changing workflows is high risk. So senior leadership sets a digital vision, and frontline staff may be eager to try new tools, but the middle freezes, and adoption stops.

This insight has shaped how The Digital Insurer, which Terry founded and operates from Singapore, approaches its work with insurers globally.

His TDI Academy programs currently engage over 50,000 insurance professionals and deliberately target that frozen middle. The GenAI Explorer program runs for 10 weeks, not two days, because behavior change requires sustained engagement rather than event-based awareness.

Participants do more than just learn about AI in the abstract. They develop personal action plans for their own business context, apply concepts through practical challenges, and build peer support networks that create accountability beyond the classroom.

Terry’s perspective is informed by operational, consulting, and entrepreneurial roles across multiple continents. As an actuary by training who has watched digital transformation unfold across Asia Pacific, Europe, the Middle East, and the Americas, he’s seen meaningful variation in both digital maturity and AI readiness — and learned that the two aren’t the same thing.

The Insurance Lead spoke with Terry about why Asia Pacific remains the most dynamic region for digital innovation, what executives consistently underestimate about AI enablement, and why effective AI adoption depends less on technology selection than on organizational readiness to actually use it.

The Insurance Lead: You’ve been working with insurers around the world for the last decade on digital transformation. What are the most significant differences you’re seeing between regions or markets in terms of digital maturity and AI readiness?

Hugh Terry: Having worked across Asia, Europe, the Middle East, and the Americas over the past 12-plus years through TDI, I see meaningful variation in both digital maturity and AI readiness.

Asia Pacific remains the most dynamic region. Markets like Singapore, Hong Kong, and parts of Southeast Asia have strong regulatory support for innovation, a consumer base that is mobile-first, and a vibrant insurtech ecosystem. Countries like China led early in digital distribution through platforms such as WeChat and Alipay, while markets like India and Indonesia are now making rapid strides, fuelled by large young populations and increasing smartphone penetration. Importantly, many Asian insurers didn’t carry the same legacy technology burden as their Western counterparts, which allowed them to leapfrog in areas like digital onboarding and claims automation.

Europe is a mixed picture. The UK and Nordics are strong on data analytics and open insurance thinking, but continental Europe can be slower due to regulatory complexity and more conservative organizational cultures. That said, GDPR has inadvertently created a sophisticated understanding of data governance that is now an advantage when it comes to responsible AI deployment.

The Middle East is emerging fast. Markets like the UAE and Saudi Arabia are investing heavily in digital infrastructure, partly driven by national-level mandates around economic diversification. We’ve seen real appetite for AI among insurers in the Gulf, though the talent base is still developing.

North America has enormous resources and some of the world’s most advanced AI capabilities, but adoption within insurance specifically can be surprisingly uneven. Many large US carriers still grapple with decades of legacy systems and fragmented data, which slows practical AI implementation despite significant investment.

The common thread globally is that digital maturity and AI readiness are not the same thing. A company can have a beautiful digital front end and still be nowhere near ready for AI because their data architecture, governance, and people capabilities aren’t there yet. That’s a distinction I find myself making frequently.

When you’re working with insurance enterprises on AI enablement, what’s the gap between what executives think they need to do versus what actually needs to happen for successful adoption?

Hugh Terry: The biggest gap I see is between executive enthusiasm for AI outcomes and the organizational groundwork required to actually achieve them. Executives tend to focus on the destination — reduced claims costs, faster underwriting, better customer experience — without fully appreciating the path to getting there.

The first gap is around people. Many executives treat AI as a technology project. They’ll approve a budget for a platform or a vendor partnership, but they underinvest in building AI literacy and enthusiasm across the organization. Successful AI adoption requires middle managers and frontline teams to understand what AI can and can’t do, to trust it, and to change how they work.

That human element is often the bottleneck, not the technology itself. This is precisely why we built TDI Academy programs the way we did — because we’ve seen repeatedly that adoption stalls without investment in people.

The second gap is data readiness. Executives assume their data is in better shape than it is. In reality, most insurers have data scattered across legacy systems, held in different formats, with inconsistent quality. Before you can run a meaningful AI initiative, you need to invest in data architecture and governance. It’s not glamorous, but it’s really necessary.

The third gap is around governance and responsible use. With AI, particularly GenAI, there are real issues around intellectual property, data privacy, and regulatory compliance that need careful consideration. Executives sometimes want to move fast without fully addressing these, which can create significant risk.

Finally, there’s a gap between pilot and scale. Many insurers have successfully run AI proofs of concept, but struggle to move from a successful pilot to enterprise-wide deployment. That requires change management, process redesign, investment in new skills and sustained leadership commitment — not just a technology refresh.

Through TDI’s programs, you’re building awareness around digital strategies and AI capabilities. What are the most common misconceptions or knowledge gaps you encounter — even among technology-focused insurance leaders?

Hugh Terry: Among the most persistent misconceptions is that digital transformation is primarily about technology. After working with thousands of insurance professionals through TDI Academy — and our broader community of over 50,000 members — I can tell you that the technology is rarely the hardest part. The challenge is changing how people think, how organizations are structured, and how decisions get made. We built TDI around a “Hearts and Minds” philosophy precisely because engagement and mindset shift are the real levers of change.

Another misconception, particularly around GenAI, is that it’s a silver bullet. I encounter leaders who believe that deploying a large language model or a Copilot tool will immediately transform productivity. In reality, GenAI is powerful but requires thoughtful implementation. You need to define clear use cases, train people on effective prompting, address data security concerns, and create governance frameworks. Our GenAI Explorer program with Microsoft exists because we saw this gap — people need structured, insurance-specific guidance on how to actually use these tools effectively.

And then there’s leadership’s misconception of the full scope of digital insurance. Many focus narrowly on one area — perhaps distribution or claims — without seeing how digital technologies affect the entire value chain from product design through underwriting, distribution, servicing, and claims. Our Digital Leadership Curriculum covers the full spectrum across seven courses and 56 lessons for exactly this reason. You can’t make good strategic decisions about AI or digital investment if you only understand one part of the picture.

There’s a false notion that digital transformation has a finish line. Leaders sometimes think they can run a program, tick a box, and move on. In reality, the pace of technological change means continuous learning is essential. The companies getting the most value from our programs are the ones that treat digital literacy as an ongoing capability, not a one-off event.

There’s a lot of pressure on insurers to “do something with AI” right now. How do you help organizations distinguish between meaningful AI initiatives and what’s essentially just checking a box?

Hugh Terry: This is one of the most important conversations we have with clients. The pressure to demonstrate AI activity is intense — from boards, investors, and competitors — and it can lead to what we’d call “innovation theater”: visible AI projects that look impressive in a press release but don’t create real business value.

We help organizations apply a simple but effective filter. A meaningful AI initiative should be tied to a specific business outcome that matters — whether that’s measurably reducing claims leakage, improving customer retention, accelerating underwriting decisions, or enabling agents to serve customers more effectively. If the initiative can’t clearly articulate what business problem it solves and how success will be measured, that’s a warning sign.

We also encourage organizations to start with use cases where AI can augment human capability rather than trying to replace entire processes. In insurance, the highest-value early applications tend to be where AI helps people make better decisions faster — claims handlers spotting fraud patterns, underwriters processing information more quickly, customer service teams resolving enquiries more efficiently. These are meaningful because they deliver tangible benefits while building organizational confidence and capability.

The difference between box-ticking and meaningful adoption ultimately comes down to whether the organization is building genuine capability or just buying technology. Technology without capability rarely works. Our job at TDI is to help build capability and enthusiasm.

When an insurer comes to you saying they want to implement AI or GenAI, what are the first questions you ask them to determine if they’re actually ready?

Hugh Terry: We typically start with five key questions that quickly reveal where an organization really stands:

• “What specific business problem are you trying to solve?” If the answer is vague — “we want to be more innovative” or “our competitors are doing it” — that tells us the strategy work hasn’t been done yet. AI should be in service of clear business objectives, not the other way around.
• “What does your data landscape look like?” We want to understand whether their data is accessible, clean, and governed properly. Many insurers have years of valuable data locked in legacy systems with no clear data strategy. If they can’t access and trust their own data, they’re not ready for AI — they need to address fundamentals first.
• “Who is sponsoring this and how broad is the support?” AI adoption that’s driven by a single enthusiastic CTO or innovation team but doesn’t have broader leadership buy-in will stall. We want to see that the CEO, business line leaders, and operations are engaged, not just IT or innovation.
• “What’s your approach to people development?” This is where our conversation often gets most interesting. If an organization is planning to invest significantly in AI technology but has no plan for engaging and upskilling their workforce, that’s a red flag. The technology will only deliver value if people understand how to work with it, and want to. This is where TDI Academy adds most value — we help bridge the people-capability and enthusiasm gaps that make or break AI adoption.
• “How are you thinking about governance and risk?” GenAI in particular raises important questions around data privacy, intellectual property, regulatory compliance, and output accuracy. An insurer that hasn’t thought about governance frameworks for AI use isn’t ready to scale it, even if the technology works beautifully in a sandbox.

These questions aren’t designed to discourage anyone — they’re designed to ensure that when an insurer does move forward, they’re set up for success rather than an expensive learning experience.

You mentioned driving enterprise adoption — not just awareness. What’s the difference, and why do so many digital initiatives stall between “we understand this” and “we’re actually doing this?”

Hugh Terry: Awareness is relatively easy to achieve. You can run a conference, share a report, or host a webinar, and people walk away informed. But information alone doesn’t change behaviour.

The insurance industry has no shortage of awareness about digital and AI; what it has is an adoption gap. Initiatives stall for several interconnected reasons.

There’s a layer I call the “frozen middle.” Senior leadership may set a digital vision and front-line staff may be eager to try new tools, but middle managers — who actually control how work gets done day to day — are often the least engaged in transformation. They’re busy, they’re measured on current performance, and changing how their teams work feels risky. This is why our GenAI Explorer program specifically targets middle managers as a key lever for change. If you can equip and motivate middle management, you unlock the organization.

Many initiatives are designed as events rather than as a process. A two-day workshop generates excitement but doesn’t build lasting capability. Our programs are deliberately structured over extended periods — the GenAI Explorer runs over 10 weeks, our Mini-MBA equivalent runs over four to six months — because behaviour change requires sustained engagement, practice, and reinforcement.

I often notice a lack of practical application. People learn best by doing, not by watching. That’s why our programs include challenges, group projects, and use case development. Participants apply what they learn to their own business context. By the time they complete a GenAI Explorer program, they haven’t just learned about AI — they’ve already started identifying and planning real applications.

Finally, adoption requires community and peer support. Change is easier when you’re not doing it alone. We deliberately build private communities within our programs so participants can share experiences, challenges, and solutions with peers in similar roles and organizations. That peer learning and accountability drives adoption far more effectively than any amount of content delivery.

Are you seeing different adoption patterns based on company size? Do smaller insurers or MGAs have any advantages in AI adoption compared to large carriers with legacy infrastructure? What about differences by line of business?

Hugh Terry: Absolutely, and the patterns are more nuanced than the simple “agile startup versus slow incumbent” narrative.

Smaller insurers and MGAs do have genuine advantages. They typically have less legacy technology debt, simpler organizational structures, faster decision-making, and a culture that’s more open to experimentation. An MGA can decide to implement a GenAI tool for underwriting analysis and have it in production within weeks. For a large carrier, the same decision might take months just to get through internal governance.

However, large carriers have advantages that are often underappreciated. They have enormous volumes of data, which is, of course, the fuel for AI. They have deeper pockets for investment. They have established customer relationships and distribution networks that provide context for AI applications. And increasingly, the large carriers that have invested in cloud migration and modern data architecture are catching up quickly. Some of the most impressive AI implementations I’ve seen have come from large insurers who’ve done the foundational work over the past five years.

The real differentiator isn’t size; it’s organizational mindset and capability. We’ve worked with large insurers like Zurich and Prudential through TDI Academy who are moving decisively on digital and AI because they’ve invested in building capability and team engagement at scale. Equally, we’ve seen smaller insurers that talk a good game about innovation but haven’t actually changed how they operate.

In terms of lines of business, we’re seeing the fastest adoption in areas where data is richest and decisions are most frequent — personal lines, health insurance, and commercial SME. Motor and home insurance have been early movers in claims automation and pricing.

Health insurance is fascinating because of the intersection of AI, wearables, and wellness data. Commercial specialty lines are catching up, particularly in areas like underwriting augmentation where AI can help process complex information faster.

The lagging areas tend to be those with less structured data or higher regulatory complexity — parts of life insurance and large commercial risk. But that’s changing as AI tools, particularly GenAI, become better at handling unstructured data like policy documents, medical records, and legal text.

Based on what you’re seeing globally, where are the biggest opportunities for AI to create actual business impact — not just operational efficiency, but competitive advantage?

Hugh Terry: Operational efficiency gets all the headlines — automating claims, streamlining underwriting, reducing manual processes — and it matters. But the truly transformative opportunities are where AI changes what an insurer can do, not just how efficiently they do it.

Hyper-personalization of the customer experience represents a major opportunity.

Insurance has traditionally been a one-size-fits-many industry. AI enables insurers to understand individual customers deeply — their behaviours, preferences, risk profiles, and life stages — and offer genuinely tailored products, pricing, and service. The insurers who get this right will build customer relationships that are very difficult for competitors to replicate. It moves insurance from a grudge purchase to a valued partnership.

Risk prediction and prevention, rather than just risk transfer, are opportunities to pay attention to as well. AI enables insurers to identify risks before they materialise and help customers avoid losses altogether. This is already happening in commercial property insurance with IoT sensors and predictive analytics, in health insurance with wellness programs powered by AI, and in motor insurance with telematics. The insurer that becomes your risk prevention partner — not just your claims payer — fundamentally changes its value proposition.

A third opportunity, and one that excites us most, is in enabling new products and markets that weren’t previously viable. AI makes it possible to underwrite risks that were previously too complex or too small to assess economically. Embedded insurance, parametric products, and on-demand coverage all become more feasible when AI can handle the complexity. This is where competitive advantage truly lies — not in doing old things faster but in doing things that weren’t possible before.

Distribution intelligence is a fourth opportunity we’ve identified. AI can dramatically improve how insurers identify prospects, match products to needs, and support intermediaries. For an industry that still relies heavily on agents and brokers, giving those distributors AI-powered tools is a powerful competitive move. Our GenAI curriculum covers AI in distribution specifically because we see enormous untapped potential here.

These opportunities require investment in people and capability, not just technology. That’s the consistent message: AI is an enabler, but the competitive advantage comes from how well your organization can use it.

Five years from now, what will separate the insurance companies that successfully leveraged AI from those that just talked about it?

Hugh Terry: Five years from now, the difference will be visible and stark. The winners won’t necessarily be the ones who spent the most on AI technology — they’ll be the ones who built the deepest organizational capability around it.

The first separator will be culture. Companies that succeeded will have embedded a culture of continuous learning and experimentation with new tech. AI isn’t a project that finishes — it’s a capability that can scale over time.

The organizations that treated AI adoption as an ongoing journey of capability-building, upskilling, and adaptation will have pulled far ahead. Those that treated it as a series of distinct technology projects will find themselves constantly starting over.

Talent is another critical separator. In the insurance arena, the war for capturing AI-specific talent in insurance will intensify, but the real winners will be those who invested in engaging and upskilling their existing workforce — the underwriters, claims professionals, actuaries, and distribution leaders who understand the insurance business deeply and can now leverage AI effectively.

That blend of domain expertise and AI fluency is incredibly powerful and very hard for competitors to replicate quickly. This is what drives everything we do at TDI Academy — building that blend of insurance knowledge, enthusiasm, and digital capability.

The use of data as a strategic asset is another important factor. Successful companies will have invested in making their data accessible, clean, and governed. They’ll have built data ecosystems that enable AI models to learn and improve continuously. Companies that didn’t do this groundwork will still be struggling with pilot projects that can’t scale.

Next is responsible AI governance. As regulation catches up — and it will — companies that built governance frameworks early will have a significant trust advantage with customers, regulators, and partners. Those who moved too fast and loose with AI will face reputational and regulatory challenges.

Finally, the fifth separator will be the customer relationship. Companies that used AI to genuinely improve outcomes for customers — better products, fairer pricing, faster service, risk prevention — will have earned the loyalty and trust that creates a durable competitive moat. Those that used AI primarily to cut costs internally will find they’re competing on price in an increasingly commoditised market.

Ultimately, the companies that win will be the ones that understood from the start that AI is a people challenge, not a technology challenge. It’s about investing in your workforce, your culture, and your customers. That’s been the core of what we’ve been advocating at TDI since we started, and I believe the next five years will prove it decisively.

Summary: How Insurers Can Successfully Drive AI Adoption

AI adoption in insurance continues to be a top priority for organizations globally, but many initiatives fail to move beyond pilot stages. As Hugh Terry highlights, the challenge is rarely the technology itself — it is organizational readiness.

To successfully scale AI in insurance, companies must focus on:

• Aligning people, not just technology
  AI adoption depends on engaging middle management and enabling teams to change how work is done.
• Investing in data readiness and governance
  Clean, accessible, and well-governed data is critical for any meaningful AI initiative.
• Bridging the gap between pilot and production
  Moving from proof of concept to enterprise deployment requires change management, process redesign, and sustained leadership commitment.
• Avoiding “innovation theater”
  AI initiatives should be tied to clear business outcomes, not just experimentation or visibility.
• Building long-term capability, not one-off projects
  Organizations that succeed treat AI as an ongoing capability, supported by continuous learning and workforce development.

Ultimately, successful AI adoption in insurance is not about implementing new tools — it is about transforming how organizations operate. Companies that invest in people, culture, and data will be best positioned to turn AI into a true competitive advantage.