Simon Glairy is a recognized expert in the fields of insurance and Insurtech, with a specialized focus on risk management and AI-driven risk assessment. For years, he has analyzed the industry’s tectonic shifts, from the adoption of smart-home technology to the strategic deployment of artificial intelligence in underwriting and compliance. Today, he joins us to dissect the latest developments, exploring how proactive prevention is reshaping the insurer-customer relationship, how massive acquisitions are redefining growth strategies, and how AI is moving from a buzzword to an essential operational tool across the entire insurance value chain.
Nationwide plans to provide 500,000 homes with Ting smart sensors for fire prevention, covering up to $1,000 in repairs. Beyond reducing claims, how does this proactive approach change the insurer-homeowner relationship? Please share some examples or metrics demonstrating this shift.
It’s a fundamental transformation from a reactive, transactional model to a proactive, continuous partnership. For generations, the relationship was dormant until a crisis occurred—a fire, a flood. Now, with initiatives like the Ting program, the insurer becomes a daily, silent guardian of the home. Imagine the peace of mind a homeowner feels, not just from a policy document, but from a tangible service that has already prevented over 1,100 electrical fires. The relationship becomes about shared interest in prevention rather than just financial indemnification. When a provider covers up to $1,000 in repairs for a hazard before it becomes a catastrophe, they’re not just an insurer; they are an active risk management partner invested in the family’s well-being. This creates a powerful sense of loyalty and trust that a traditional policy simply cannot replicate.
AIG recently completed a $2.7 billion acquisition of stakes in specialty insurer Convex and asset manager Onex. How do these dual investments in underwriting and asset management create a synergistic strategy for growth, and what is the expected impact on AIG’s earnings and equity?
This is a masterclass in strategic capital allocation that creates a powerful flywheel for growth. By taking a significant stake in a specialty insurer like Convex, AIG is directly tapping into a high-margin underwriting business. Simultaneously, investing in a sophisticated asset manager like Onex bolsters the other side of the insurance balance sheet—the investment portfolio. The synergy is clear: AIG can leverage Onex’s expertise to manage its own assets more effectively while participating in the underwriting profits from Convex. This dual approach diversifies AIG’s revenue streams, making it less dependent on the volatility of its own underwriting cycles. The expectation, as stated, is that this will be accretive to earnings and return on equity starting in 2026, creating a more resilient and profitable enterprise in the long run.
Insurtech platforms like TheZebra.com are appointing Chief AI Officers with backgrounds from major tech firms like Google. What specific, practical changes does this leadership bring to an insurer’s engineering and customer-facing tools? Provide a step-by-step example of a new AI implementation.
Bringing in a leader from a place like Google injects a completely different DNA into the organization. It’s less about simply buying AI tools and more about building a culture of deep, integrated intelligence. The immediate change is a shift from siloed data projects to a unified AI infrastructure, much like the one used to train massive models like Gemini. Let’s walk through an example for a customer-facing tool. Step one: The new CAIO implements a system to unify all customer interaction data—clicks, quotes started, questions asked, call transcripts. Step two: They use this data to train a proprietary model that understands user intent and confusion points in real time. Step three: This model is deployed as an “advisor enablement” tool. When a customer is struggling on the website, instead of a generic pop-up, the AI flags the specific point of friction—say, confusion over liability limits—and provides a human agent with a tailored script and a simplified explanation to proactively assist that customer. This isn’t just about faster quotes; it’s about creating a hyper-personalized, empathetic journey that anticipates needs before the customer even has to ask.
The partnership between AI-native carrier MGT and wholesale distributor Amwins aims to reduce underwriting for some E&S risks from days to instants. What are the biggest operational hurdles in deploying this model at scale, and how does it change the role of the traditional underwriter?
The primary hurdle is data integrity and integration at scale. For an AI to make an “instant” decision on a complex E&S risk, it needs a massive, clean, and continuous flow of data—property records, business financials, location-specific risks, and more. Integrating MGT’s sophisticated AI platform with Amwins’ vast distribution network, which places $45 billion in annual premiums, is a monumental IT and process engineering challenge. You have to ensure the data pipelines are robust and the AI’s decision-making logic aligns with the real-world nuances that experienced brokers understand. This profoundly changes the underwriter’s role. They are no longer data gatherers or manual processors. Instead, they become portfolio managers and exception handlers. Their expertise is redirected toward the most complex, unusual risks that fall outside the AI’s confidence threshold, and they spend their time refining the AI models, analyzing portfolio performance, and building strategic relationships, which is a much higher-value use of their talent.
Logic Underwriters is adopting AI storm models in Texas to analyze property-specific details like roof condition and yard debris. How does this granular, AI-powered risk assessment compare to traditional methods, and what are the trade-offs for insurers when implementing such precise underwriting technology?
It’s the difference between using a regional weather map and having a dedicated satellite over every single property. Traditional methods rely on broad geographical zones—a ZIP code, for instance—to price risk. This is a blunt instrument that often misprices properties, penalizing well-maintained homes in a high-risk area and underpricing poorly maintained ones. AI-powered models, like those from ZestyAI, zoom in to the micro-level, analyzing aerial imagery to assess specific roof conditions, the risk posed by overhanging vegetation, or even the presence of yard debris that could become a projectile in a high wind event. This granular approach allows for incredibly precise, fair pricing. The primary trade-off is the initial investment in technology and data, as well as the need for regulatory approval, which ZestyAI has in 30 states. Insurers also have to manage the potential for customer pushback if a hyper-specific analysis leads to a higher premium, which requires a new level of transparency in communicating risk factors to policyholders.
New compliance platforms like ZeroDrift are using AI to check communications against regulations in real time before they are sent. Can you describe the technical process for encoding complex SEC and FINRA rules into a system that can provide instant feedback on an email or post?
It’s a fascinating multi-step process that blends legal expertise with advanced natural language processing. First, the platform ingests the raw regulatory texts from sources like the SEC and FINRA, along with a firm’s own internal policies. These are dense, legalistic documents. The second step is crucial: these rules are broken down by legal and AI experts into “machine-readable rulepacks.” This involves translating ambiguous legal language into concrete, logical conditions that an algorithm can understand—for example, a rule against making “promissory statements” is translated into patterns of words, phrases, and sentiment that the AI can detect. Third, the system uses natural language understanding (NLU) models to analyze the context, sentiment, and specific claims within a piece of communication, like an email draft. Finally, as the employee types, the platform’s API checks the content against the rulepacks in real-time. If it detects a potential violation, it doesn’t just block it; it flags the specific issue and suggests an compliant alternative, effectively acting as a compliance co-pilot at the very point of creation.
With INSTANDA and ServiceNow partnering to integrate no-code policy administration with AI-driven workflows, what are the primary benefits for insurers looking to modernize legacy systems? Please walk through how this integration could improve the claims process from start to finish.
The primary benefit is a massive leap in both agility and efficiency, freeing insurers from the shackles of their old, rigid legacy systems. No-code platforms like INSTANDA allow business users to design and launch new insurance products in weeks instead of years. When you combine that with ServiceNow’s powerful AI-driven workflow engine, you create an end-to-end intelligent ecosystem. Let’s trace a claim. A policyholder initiates a First Notice of Loss through a mobile app. The ServiceNow AI immediately triages the claim, using data to determine its complexity and severity. Instead of a human manually assigning it, the system automatically routes a simple claim—say, a cracked windshield—to a straight-through processing workflow for instant approval and payment. A more complex claim, like a multi-vehicle accident, is intelligently routed to a specialized human adjuster, with all relevant policy information, photos, and a preliminary AI damage assessment already attached. This integration transforms the claims process from a linear, manual sequence into a dynamic, intelligent, and dramatically faster experience for both the customer and the insurer.
What is your forecast for insurtech in the next 18 months?
Over the next 18 months, I predict we will see a significant “flight to quality” and consolidation in the insurtech space. The era of speculative investment in flashy apps is over; the focus now is on technologies that deliver tangible, immediate ROI. We’ll see three key trends accelerate. First, embedded insurance will become the default, with proactive risk-prevention services, like Nationwide’s Ting program, being bundled directly into policies as a standard offering, not a novelty. Second, generative AI will move from pilot programs to full-scale deployment in core operations, particularly in underwriting for complex commercial and E&S lines, turning what was a week-long process into a minutes-long one. Finally, the strategic partnerships between AI-native carriers and established giants, like the MGT-Amwins deal, will become the dominant model for innovation, as it allows for the perfect marriage of cutting-edge technology with massive distribution scale. The winners will be the companies that can seamlessly integrate these advanced tools to not just manage risk, but to actively prevent it.
