What’s at stake if insurers rush into AI too early?

Insurers may be in a race to adopt and scale artificial intelligence faster than their competitors do, but Norm Hudson, CEO of Staff Boom, cautions that jumping the gun could lead to disaster.

Hudson told InsuranceNewsNet that AI pilots launched without a vision of how to handle different scenarios, prerequisite basics like a standard operating procedure and adequate testing are destined to fail.

“Setting the foundation is critically important to anything automation-related, or even within a manual business process,” Hudson said. “Having your standard operating procedures, knowing how you will deliver and deal with the various scenarios that come up, is critically important.”

He said in his experience, businesses can “find themselves in a spot where they don’t have naming standards or operational consistencies,” but layer automated or AI processes on top and run a “narrow pilot” in a controlled setting with a focus.

While AI systems can appear to function well in controlled pilots, Hudson explained that the proof is revealed in the pudding once that “narrow pilot” is moved out into the real world.

“Those variables crash the test. There are too many things for it to come through, it confuses the model and the work doesn’t get done because the foundation is not set right. That is a common problem right now,” Hudson explained.

Without the proper foundational data, an AI system “is not set up in a manner that will allow it to evolve and have success in a variable environment.”

Change management exacerbated

An AI tool struggling to function effectively with an unstable foundation creates a domino effect, Hudson noted. Employees then refuse to use the AI tool, effectively ending the AI pilot’s life cycle long before it’s implemented at scale.

“You start introducing all the variables, and the systems aren’t working properly, and then separate. Apart to that is the human element, where people have been slow,” Hudson said.

“They don’t want to self-determine themselves, and so you see people who are not using the tools. That has been the consistent feedback with the AI-focused providers out there - that even when they have the resource provided for them, they see low utilization.”

Hudson noted that change management is already a challenge on its own, as many humans naturally tend to resist change, and some employees are particularly resistant to using AI. However, he added that trying to get employees to adopt an already flawed system built on an inadequate foundation exacerbates that.

“To the extent that you roll out a pilot and it’s not running smoothly, all they’re going to see in that is the flaws. A lot of times, the sentiment there isn’t, ‘Well, how do I lean in to make this better?’ It’s, ‘Oh, it doesn’t work, and I want to pivot and walk away.’ We’re certainly seeing that in the space in and around AI and some of its complexity,” he said.

He noted that the insurance industry is at particular risk for this because its aging workforce means most employees are “well into their 50s, 60s and 70s and not as in-tune to technology as the 25- to 30-year-old group would be.”

“So as you see the variables in the work product, that becomes the benchmark of what it is. They don’t see any of the good. They focus, in a lot of cases, on what’s not working, perfect or optimized, and when that’s your lens, it’s really easy to tank a pilot,” Hudson said.

Not beyond redemption

The good news, Hudson said, is even if an AI pilot fails, companies can still “come back and make another run at it.”

“I don’t think anything is so far gone that it can’t be revisited. It’s a function of, OK, what did we learn from it? How do we reapproach this with a fresh perspective and increase our chance of having a successful pilot rollout in the next iteration? I think that’s commonly what’s taking place,” he said.

However, he also suggested how insurers can avoid a failed pilot by doing the work upfront instead of rushing to roll out too soon.

“The biggest thing is lean in, set your operating procedures, get your data organized, set your foundation right, because that will give you a massive increase in the probability of success,” Hudson said.

“And then, as you’re running through your pilot, don’t run it so narrowly that you’re not getting a real-world example of what’s ahead. I think just being aware of the challenges ahead and tabling this where you have a broad enough funnel that you’re getting a real-world understanding of the challenge that you’re solving for is really important.”

Staff Boom is a private insurance business process outsourcing company founded in 2014 and based in Anaheim, Calif.

© Entire contents copyright 2026 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.

Rayne Morgan is a journalist, copywriter, and editor with over 10 years' combined experience in digital content and print media. You can reach her at [email protected].