by Lee Boyd, President, Claims Operations, Alacrity Solutions
You can’t attend an insurance conference or open your inbox without hearing about artificial intelligence (AI). What was the wave of the future is now the present. Regardless of your perspective on it, AI will define the evolution of many industries in the years to come, ours included.
But what exactly does it mean for insurance? And more importantly: what do we do with it?
I’ve spent two decades at the intersection of insurance and technology, including running claims operations through periods of both technological stagnation and transformation. Buzzwords, platforms, and big promises can make tech conversations feel disconnected from real claims work. But for those who’ve spent years in the trenches, certain truths become clear.
Most tech investments fail because no one owns the problem.
For one, nobody tells you how often tech investments fail, not because the tools didn’t work, but because no one owned the problem they were meant to address. That kind of failure isn’t about the tech. It’s about the absence of strategy. And avoiding it starts with clearly defining your problem.
Defining The Problem Comes First
Before any system is implemented, insurers must ask a fundamental question: what problem are we looking to overcome? If you want to implement AI because it’s trendy, you’re going to have a hard time finding value.
Success starts with three things:
• A clear problem definition
• A defined owner
• A shared vision for success
What are we trying to win? What are we willing to lose? Initiatives fail without this foundational clarity. Tools rushed in without a plan, or force-fit into a part of the organization that didn’t ask for it, often fail.
AI As A System, Not A Task
AI doesn’t work well when it’s bolted onto the end of a process. For AI to deliver meaningful results, it has to be designed into the flow, from the start.
That’s when it becomes additive: keeping the full picture of the claim in view, flagging issues a human might miss, and supporting faster, more accurate decisions. We’re already seeing promising examples in claims triage, subrogation, and fraud detection.
Take ClearSpeed, which uses voice AI during FNOL to flag potentially suspicious claims. Or Charlie AI, which is building models to help carriers spot fraud earlier. These tools are embedded at critical points in the claims process. That’s where the value lives.
Of course, embedding AI early doesn’t mean blindly trusting it. That brings us to one of the most persistent barriers in the industry: trust.
Trusting What You Can’t See
The trust deficit makes sense. How can decision-makers, especially in a historically risk-averse industry, feel confident in a system they can’t fully see?
Using AI without guidance creates a black box. You’re not sure what’s happening between input and outcome.
There are two ways to fix this:
• Educate yourself. Know what’s possible. Read beyond your comfort zone. Talk to peers. Especially those who’ve failed.
• Design intentionally. Build frameworks (rules, use cases, oversight) that guide how the system works and what it should (and shouldn’t) do.
With structure in place, AI stops being mysterious and starts becoming measurable.
Culture Moves Slower Than Code
In many cases, the biggest obstacle isn’t technology, but inertia. Insurance is a methodical industry. But we’re also sitting on an enormous stockpile of data. AI thrives on data. That makes insurance a perfect use case, but only if we’re willing to move.
I’ve had carriers tell me, “I love what you’re doing! Just don’t do it with my stuff.” That mindset is understandable, but risky. Waiting to “fast follow” isn’t a viable strategy anymore. There are doers, and there are those who get left behind.
Yes, the stakes are real. Budgets are tight. Leaders don’t want to fail in public. But in this landscape, inaction can be just as risky as experimentation. Most AI projects that fail do so because they lacked vision, ownership, or infrastructure, not because the technology didn’t work.
Move forward deliberately. Start small. But start.
AI Is A Commitment, Not A Project
AI isn’t a magic button. It’s an operating model shift.One that only works when leaders approach it with clear goals, tight controls, and a long-term view. The carriers who win with AI won’t be the ones who move the fastest—they’ll be the ones who move with purpose.
AI isn’t a magic button. It’s an operating model shift.
When implemented with that level of intent, AI doesn’t just improve speed or cost efficiency. It unlocks higher-order outcomes: better decision-making, more consistent service, and new space for human involvement to shine. Those are the conversations we need to start having next.
Lee Boyd is is a veteran executive with over two decades of experience in the insurance industry. He currently serves as President of Claims Operations at Alacrity Solutions, where he leads a large-scale claim service organization supporting carriers, MGAs, and businesses across the property and casualty industry. His teams specialize in managing complex claims operations across field adjusting, desk adjusting, third-party administration, subrogation, call center operations and claims support services. Lee is also the host of FNO: InsureTech Podcast, where he interviews founders, operators, and industry leaders about how technology is reshaping the insurance industry.
Alacrity Solutions is one of the largest independent providers of insurance claims management services in North America, delivering property, auto, and casualty claims adjustment, staffing support, temporary housing services, plus subrogation and recovery solutions. With more than 50 years of experience, Alacrity combines skilled adjusters with data intelligence to deliver faster, more accurate, and more consistent outcomes. For more information, please visit www.alacritysolutions.com.