The insurance industry, historically built on meticulous risk assessment, is undergoing a transformation. With the integration of artificial intelligence (AI) and third-party data, insurers now have tools capable of reshaping how they quantify, model, and manage risk. During a recent Unstructured Unlocked podcast episode, Tom Wilde, CEO of Indico Data, and Michelle Gouveia, VP at Sandbox Insurtech Ventures, explored these shifts in an engaging conversation that highlighted key developments and opportunities within the field.
This blog takes a closer look at their discussion, covering the unique potential AI brings to the table, the opportunities presented by third-party data, and the actionable insights needed to adapt to these advancements effectively.
Listen to the full podcast episode here: Unstructured Unlocked Podcast
Unlocking smarter data with AI and insurance
One of the most significant opportunities in the insurance industry today lies in leveraging third-party data for everything from fine-tuning claims processes to improving underwriting accuracy. However, this transition is not without challenges. Tom Wilde pointed out that the core of insurance has always been about probabilities. From actuaries determining risk to developing precise catastrophe models, the foundation of the industry depends on aggregating and analyzing the right data.
The inclusion of third-party data sources, such as satellite imagery, geospatial tools, and even real-time IoT sensors, offers insurers a richer understanding of risk. These data tools help:
- Paint a more accurate picture regarding natural disasters and environmental hazards
- Simplify the assessment of complex multi-property insurance portfolios
- Build more robust risk models tailored to client needs
Yet, Wilde highlights a critical distinction between data access and data reliability. To confidently incorporate these tools, insurers must validate their accuracy and ensure their integration aligns with enterprise-grade governance standards.
Generative AI and the rise of agent intelligence
Generative AI and agent intelligence (or “agentic AI”) are among the most discussed technological advancements disrupting traditional insurance workflows. But these concepts often lead to debates about readiness and execution.
Tom Wilde described agent intelligence as involving orchestrated AI systems designed to assist in actionable workflows. For instance, in underwriting, generative AI might suggest a “next best action,” such as requesting additional information from brokers or automating the integration of third-party risk data into existing processes. These intelligent agents aren’t replacing humans; instead, they are streamlining repetitive tasks, enabling underwriters and claims adjusters to focus on higher-value responsibilities.
Wilde noted two approaches insurers are taking with generative and agent AI adoption:
1. The high-accuracy path
This approach prioritizes building highly accurate, well-tuned AI models that perform consistently and predictably. While effective, it requires significant upfront resources for accuracy calibration and time to deployment.
2. The good-enough path
Here, the focus is on deploying AI systems quickly to achieve initial gains in efficiency or triage capabilities, with less emphasis on perfect accuracy. These systems act as first-line filters, routing complex cases to experienced underwriters or adjusters for deeper review.
Both strategies have their merits. What they share, however, is the necessity for robust oversight, particularly when expanding AI’s role into more decision-critical insurance applications.
The buy vs. build debate in AI adoption
A recurring question for insurers grappling with AI and automation is whether to develop proprietary AI tools internally or partner with third-party vendors. Wilde succinctly framed this debate using an analogy involving saws, carpenters, and architects. He explained that while AI models serve as powerful tools (like a saw), the challenge lies in whether insurers see themselves as capable of building a substantial technological infrastructure (like a skilled carpenter) or designing an enterprise-grade system from the ground up (like an architect).
For insurers considering in-house solutions, key questions involve:
- Do we have the capability to maintain the system long-term?
- Can our internal AI system keep pace with evolving regulatory requirements, especially in regions where compliance standards are rigorous?
- Are there unique strategic advantages (e.g., better data management or faster deployment timelines) that justify this investment?
These considerations often tilt the scale toward buying instead of building. Established vendors bring years of expertise solving domain-specific challenges in claims and underwriting automation, enabling insurers to benefit from a proven, off-the-shelf solution that reduces operational strain.
Transforming underwriting and claims with AI
The integration of AI into underwriting and claims workflows has already showcased tangible benefits. Wilde highlighted examples where AI reduces cognitive load on underwriters and streamlines the flow of case files across different teams, allowing enterprises to operate with greater speed and precision.
Notable use cases include:
- Automating document processing for claims adjusters, reducing manual reviews and improving first-pass accuracy.
- Supporting risk-intensive industries, such as construction or marine insurance, by detecting patterns of probable losses earlier in the value chain.
- Increasing customer satisfaction with faster, more accurate settlements.
When deployed thoughtfully, AI simplifies compliance roadblocks while ensuring legacy systems evolve into agile ecosystems capable of handling next-gen demands.
Preparing for the AI-led insurance of tomorrow
The competitive edge for insurers over the next five years will involve more than advanced AI tools. Wilde predicts that organizations positioning themselves to thrive will master three critical areas:
1. Context and domain expertise
Insurers must bring a unique perspective to their market challenges by aligning solutions directly with client pain points. This contextual advantage will separate leaders from the pack.
2. Distribution and accessibility
Understanding distribution networks, particularly broker behavior and buying patterns, becomes critical. Those mastering end-to-end market channels are likely to win client trust over those leaning solely on technology.
3. A supercharged workforce
By focusing inward, companies can create technologically empowered teams, minimizing cumbersome workflows that detract from motivation and efficiency. Wilde notes this as a major opportunity overlooked by insurers emphasizing external strategies more than internal transformation.
While advanced AI provides an exciting future, the key lies in how enterprises systematically unlock its full potential across claims, underwriting, and customer lifecycle management.
Advancing together into the AI future
AI and third-party data are clearly reshaping the insurance landscape in fundamental ways. From groundbreaking risk analysis models to smarter underwriting systems, these technologies are enabling advancements that are redefining operational playbooks across the sector.
However, adoption isn’t about chasing every shiny new tool. It’s about pairing the right solutions with actionable goals, ensuring they align with organizational culture, and maintaining long-term sustainability in execution.
To thrive in this evolving landscape, insurance firms must balance ambition with execution, integrating these innovations not just as tools, but as collaborative extensions of their strategic vision.
Subscribe to the Unstructured Unlocked podcast to get the latest episodes on your favorite platforms, including:
