As AI adoption accelerates across insurance, many agencies, MGAs, and carriers are looking for practical guidance on what actually works. Emily McGinn, General Manager, MGA and Wholesale at Vertafore, shares straightforward advice on where to start with AI, what defines success, and how to use technology to improve underwriting workflows without losing human judgment.
Q: How are you thinking about AI in insurance?
Insurance is highly regulated and comes with significant compliance requirements. Right now, we’re using AI to optimize the decisions that humans need to make, rather than replacing those decisions. We’re focusing on intake, pattern recognition, and providing summaries, as well as reducing handoffs so we can get to the decision-maker faster.
Think about the underwriter—the person assessing the risk of a policy. We look at that workflow and ask: How do we get the right data to the underwriter as quickly as possible so he or she can make the right decision? Throughout that process, there have traditionally been a lot of manual handoffs. Our goal is to optimize and prepare the experts with the best information, without having AI drive the decisions directly, because we have to be careful given the regulatory environment.
If you look at climate risk in certain areas, it’s changing so quickly that the way policies were rated for fire risk ten years ago is completely different today. In fact, it can vary from one side of the street to the other. Those are situations where you still need an expert to evaluate what needs to change. We can’t rely on AI to make those decisions just yet.
Q: For smaller organizations in regulated industries like insurance, what do you think is the best way to begin implementing AI?
I would start with low-risk workflows. That’s table stakes.
For example, at Vertafore, streamlining submissions for MGAs has been a strong starting point for bringing AI into our solutions. These are applications collecting basic information—birth dates, names, and similar data. So intake and summarization are where we began. If something is incorrect, you can quickly go back and update a middle name or validate a few additional details.
You have to be very deliberate about where you insert human checkpoints. It’s not about inserting AI and seeing what happens. It’s about clearly defining where a human review is required and what that review should look like.
That’s why it’s critical to design the operating model first and then apply the tool. AI is a tool; it is not the operating model.
Q: What is a key lesson or a surprise you found when implementing AI within products?
As we advanced our AI initiatives, we took a disciplined approach—introducing AI in select areas, measuring impact, and refining the broader end-to-end workflow to ensure we were driving meaningful improvements. That focus on workflow optimization has been critical to maximizing value.
We had to speak with many of our customers, from individual agents to large carriers, to understand their specific pain points. Without doing that upfront work, we might have assumed we understood the pain points from our perspective. But we needed to identify where the real inefficiencies were, where people were “swivel-chairing” between systems, where time was being wasted, and where providers weren’t getting the right information when they needed it.
We ultimately determined that workflow design had to come first.
How should the distribution channel operate? What does the underwriter need to see? How do we optimize their experience? From there, we could go piece by piece and decide where AI should be used, where it shouldn’t, and what guardrails needed to be in place around each use case.
Q: How do you lay out a roadmap that avoids common mistakes and keeps timelines realistic?
Nothing is replacing humans in the workflow. We haven’t seen that, and that’s not the goal with AI. We’re focused on optimization.
Our measure of success isn’t simply how many policies an underwriter can review today versus a year ago. It’s how AI helps remove repetitive tasks so underwriters can spend more time on judgment and complex decision-making.
Once the process is optimized, the underwriter is sitting there with the right information in front of them. In theory, the call sheet or detail sheet that used to take three months to assemble can now be completed in three weeks, three days, or eventually even three hours.
The goal is to amplify each individual’s impact, not by replacing their judgment, but by removing friction and accelerating the tasks that support strong decision-making. AI allows them to apply their expertise more broadly and focus on higher-value work.
Q: What is the biggest bottleneck preventing organizations from seeing success with AI?
One of the biggest challenges we see is what we call “decision latency.” That’s the gap between when information becomes available and when a human decision is actually made. AI can dramatically speed up how quickly work moves to the next step. But if the decision-making capacity downstream doesn’t evolve along with it, work simply begins to queue up in a different place.
In other words, the bottleneck often isn’t the AI tool itself. It’s what happens next in the workflow. If AI increases daily submissions by 20%, but the surrounding processes, staffing models, or review structures aren’t aligned to handle that shift, the organization won’t fully realize the benefit.
That’s why end-to-end optimization matters. AI should reduce friction across the entire process, not just accelerate the front end. When each step is aligned, decisions happen faster, expertise is applied where it adds the most value, and the organization sees meaningful gains rather than just redistributed volume.
Q: How does AI implementation ultimately benefit the insured?
It’s not just about making companies more efficient. In the insurance industry, improving efficiency ultimately delivers better value to the end customer.
When processes are optimized and data is clearer, insurers can become more precise in how they assess and assign risk. Instead of “peanut butter spreading” risk broadly, they can be much more targeted and tactical in their approach.
That level of precision benefits the end insured.
About Emily McGinn
At Vertafore Emily is the General Manager of the MGA business unit, responsible for product development, professional services and customer service for all Vertafore’s MGA and Wholesale clients.
Emily McGinn is a dynamic and accomplished business and technology executive with extensive experience spanning finance, sales operations, human resources, and digital transformation. Recognized for her collaborative and results-driven leadership, she has guided organizations through complex transformations that accelerate growth and strengthen performance.
Watch the full webinar, “The Handoff Problem: Making AI and Humans Work Together in 2026.”