BI & Send Webinar Recap Blog: Beyond the Hype - Fixing the Real Problem in Underwriting Transformation

How AI Is Rewriting the Underwriting Playbook

AI is moving quickly through the insurance sector, but underwriting teams are progressing at very different speeds. Some are experimenting, some are piloting, and a smaller group is beginning to scale. During this webinar, Dean LaPierre, CUO with CrossCover, Robbin Botnick, Country Manager with Probitas 1492, and our own Brad Tabor, Head of Revenue North America, shared where organizations are gaining traction, where they’re hesitating, and what it will take to operationalize AI responsibly and effectively. Two live audience polls added an important layer of insight concluding that most organizations are still early in their journey, and concerns around third‑party AI risk remain high. Together, the discussion and the data painted a clear picture of an industry that sees the opportunity but is still working through the practical realities of adoption

Early Momentum, Uneven Progress

Before diving into the challenges and opportunities, attendees were asked where their organizations currently sit on the AI underwriting maturity curve. The results reflected a market still in its early stages. 

  • 26.5% are exploring potential use cases
  • 22.1% are running limited pilots
  • 16.2% are scaling across multiple workflows
  • 14.7% are using AI broadly across underwriting
  • 20.6% are still unsure or defining their strategy

As one panelist noted, these numbers would have looked even more conservative six months ago. We can see that momentum is building, but the industry is far from mature.

What’s Slowing Teams Down

Much of that slow progress can be traced back to the same set of barriers. Legacy systems remain one of the heaviest anchors. Many carriers still operate across multiple platforms that don’t integrate, creating fragmented data and inconsistent workflows. As one speaker put it, “If none of the legacy systems talk to each other… that’s always going to be a barrier.” Without a unified data foundation, AI output becomes inconsistent and difficult to trust. The technical challenge is only part of the story. There is also a very human element at play. Dean described a “primal fear” that many leaders feel. The fear of the unknown, of regulatory exposure, of losing control. Insurance is a fiduciary business, and that responsibility often leads to caution, even when the potential upside is clear. It’s a dynamic that can slow decision‑making and create what he called “analysis paralysis.”

Training surfaced as another critical gap. “If we don’t invest in training, AI is going to be a real challenge for teams,” Robbin emphasized. Underwriters need to understand not just how to use AI, but how it fits into their judgment, their workflow, and their relationship with brokers.

Where AI Helps—and Where Humans Still Lead

As the conversation shifted to where AI can add the most value, the panel drew a sharp distinction between tasks that benefit from automation and those that require human expertise. AI excels at the repeatable including data extraction, appetite checks, coverage comparisons, document summarization. These tasks are rules‑based and time‑consuming, making them ideal candidates for automation. Underwriting, on the other hand, is not a mechanical discipline. It’s a blend of data, experience, and negotiation. As Dean put it, “You’re entitled to your own opinions of risk, but not your own facts.” AI can deliver the facts; underwriters interpret them. Structuring, pricing nuance, broker negotiation, the final stages of a deal, remain firmly human‑led.

This distinction becomes even more important when considering the diversity of business written across the market. Lower middle‑market risks can support higher levels of automation, while specialty risks, with their complexity and bespoke structures, require more human oversight. Technology must flex to the line of business.

The Orchestration Challenge

As carriers introduce AI into underwriting workflows, the challenge becomes one of orchestration. Without a clear strategy, organizations risk adding more tools, more data sources, and more steps, ultimately slowing teams down. The panelists emphasized the importance of a single workbench or shared data layer, a place where information flows consistently and where automation can be applied selectively. “You don’t have to automate everything,” Robbin noted. “You can automate parts of the workflow.” The goal is not full automation; it’s intelligent automation that supports underwriters without constraining them. Platform partners play a critical role here, bringing governance, security, and integration capabilities that allow carriers to focus on underwriting logic and business outcomes.

Vendor Risk Is Top of Mind

The second audience poll revealed just how much attention carriers are giving to third‑party AI and vendor risk. When asked how concerned they were about these dependencies in underwriting decision‑making, half of respondents (50%) said they were somewhat concerned, and another 23.3% said they were extremely concerned. Only 6.7% said it wasn’t a major concern, and 3.3% hadn’t yet evaluated it.

This aligns with the panel’s perspective. Data ownership was a recurring theme. “If you don’t own your data,” Brad warned, “your data is going to be used in models provided to your competitors.” Transparency into how models are built and trained is equally important. Black‑box systems are incompatible with a regulated industry where carriers remain accountable for every decision.The Path Forward

Preparing for What Comes Next

The session closed with a forward‑looking discussion about how underwriting teams can prepare for the next phase of AI adoption. The guidance was pragmatic: anchor AI initiatives to measurable business outcomes, invest in continuous training, improve data quality now, encourage experimentation, and move deliberately—because standing still is now the bigger risk. As Dean summarized, “Nothing good has really happened without embracing risk.” The organizations that thrive will be those that modernize thoughtfully, invest in their people, and adopt technology with intention.

Watch the full webinar here.

Categories:
  • Insights
Tags:
  • AI
  • webinar

Related Resources

Insights

INFUSE Recap: Why Climate Risk Now Lives in the Underwriting Workflow

Climate risk is no longer a future concern. It's already reshaping underwriting, capital, and risk decisions across the insurance industry.
INFUSE_blue-776x393-1
Read more
Read more
Insights

Trust at Scale: Governance, Auditability, and the Future of Agentic AI in Insurance

Trust at scale needs evidence. Here's how the Send Agentic Framework keeps AI decisions controlled, auditable, and regulator-ready.
Agentic Framework 3
Read more
Read more
Insights

Why underwriting transformation keeps falling short and what will fix it  

Underwriting transformation keeps disappointing. Not because the tools are wrong, but because the architecture is.
Underwriting transformation blog
Read more
Read more

Underwriting Resources

Guide

Underwriting Maturity Framework: Moving from a process-driven to a data-driven operating model

Download guide
Guide

Grow your delegated underwriting

Send_ResourceImages_2026_Grow Your Delegated Portfolio
Download guide
Insights

Is your underwriting platform really delivering business value?

Read more
Company News

Send achieves ISO 42001 certification, setting the standard for controlled agentic AI in underwriting

Read more
Trends

Top 10 insurance industry trends shaping underwriting in 2026

Read more

Ready for frictionless underwriting?