The future of AI in insurance is becoming clearer
ITC Vegas 2025 brought together thousands of leaders across the insurance ecosystem—from carriers and brokers to insurtech innovators and AI practitioners. And while the show floor was packed with demos and dashboards, the real conversation was about where the industry goes from here.
The hype around generative AI isn’t gone, but it’s no longer enough. Across sessions and side conversations, one theme kept emerging: it’s time to move past prototypes and pilot fatigue toward scalable, orchestrated, real-world impact.
Here are five of the biggest takeaways from this year’s event—through the lens of the insurance leaders who are actively driving that shift.
1. Much of the industry is stuck at “everyday AI”
Underwriting assistants, claims bots, and FAQ-style chat tools have delivered value—but in narrow, siloed ways. Across the insurance value chain, there was talk of “LLM fatigue,” i.e., a weariness of seeing the same demos, hearing the same pitches, and evaluating tools built on identical models with little differentiation.
What’s missing is orchestration, i.e., a scalable, connective layer between data, tools, and teams that allows insurers to readily build out AI use-cases that actually deliver for end-users (be they policy-holders or internal stakeholders). A major keynote framed the path forward as a three-stage AI maturity curve:
- Everyday AI: Task-specific helpers that answer or retrieve
- Domain AI: Context-aware tools that understand insurance workflows
- Agentic AI: Coordinated systems that take action across steps and systems
That’s why so many vendors are suddenly claiming “agentic AI” — using the term to rebrand narrow tools as something more. One speaker called the trend “agentwashing.”
The real innovation, however, will come from orchestration: the connective layer that allows insurers to evolve from helpers to true AI agents.
2. Yes, people are still very necessary
There was a clear tonal shift from previous shows regarding human capital. Insurance leaders are rejecting the narrative that AI must come at the cost of headcount. (This was a prevalent—if not explicit—narrative in recent years.) Instead, there's a push to rethink how AI can—and indeed, is—unlocking opportunities for high-value work.
Executives shared examples of teams being repurposed from manual claims intake to more complex resolution roles, or from repetitive data checks to personalized client outreach. Rather than automating people out of the equation, insurers are finding success building symbiotic workflows—where AI handles the repetitive grind, and humans focus on judgment, service, and strategy.
3. The industry is shifting towards centralized, efficient AI tech stacks
The bevy of new vendors, tools, and offerings is, on one hand, great—there’s plenty of innovation. But on the other hand, it’s becoming unwieldy to manage. From CIOs to operations leads, the message at ITC was consistent: “We don’t need more tools—we need better ones that do more.”
Carriers have moved beyond buying tools just because they know they need to do AI. They’re now focused on solving complex, cross-functional problems that span multiple systems and teams. That requires platforms that can connect the dots—not point solutions that handle only one step in isolation.
Instead of proof-of-concept pilots for narrow use cases, insurers are prioritizing how solutions fit into larger digital transformation efforts. Can it scale across lines of business? Does it support multiple workflows from a single interface? Can it serve both internal users and external policyholders?
The technologies that make a true difference aren’t just feature-rich—they’re adaptable, extensible, and built to handle the real-world complexity of the insurance enterprise.
4. The fastest wins are happening behind the scenes
While customer-facing automation remains a strategic priority, many of the most compelling AI use cases discussed at ITC 2025 were inward-facing. Carriers are increasingly deploying AI agents and assistants to streamline underwriting, broker enablement, policy servicing, and internal support workflows.
From reducing “swivel-chair” tasks to accelerating quote generation and internal knowledge retrieval, internal efficiency gains are becoming a primary lever for realizing near-term ROI. For many insurers, this is proving to be the easiest and most immediate on-ramp to enterprise AI adoption.
5. Compliance can no longer be an afterthought
As scrutiny of AI continues to evolve, insurers are compelled to navigate an evolving patchwork of regulations at all levels of government. When it comes to AI, explainability, audit trails, and role-based access controls are not nice-to-haves but are table stakes, especially for tools used in consequential workflows like underwriting and claims management.
The message at ITC was clear: compliance isn’t something you bolt on later. It’s part of how insurers define “enterprise-ready.” Platforms that can’t show how they work—or why they made a decision—are quickly falling off the shortlist.
Bottom line
ITC 2025 felt like a turning point. The optimism around AI is still strong—but it’s more grounded, more skeptical, and more focused on integration, outcomes, and long-term value.
For carriers and insurance leaders, the question is no longer if AI matters. It’s:
Where is it working today? What’s ready to scale? And how do we move from pilots to impact?
Ready to see what agentic AI looks like in action?
of how DRUID orchestrates AI agents across claims, underwriting, service, and more—no black boxes, no guesswork.
