Jess Mand over at Insurance Journal reached out to me a few weeks ago to talk about AI, carriers, and the governance gap that's growing between them. The piece just dropped. It's called The Shadow AI Problem in Insurance: Navigating Trust, Oversight and Innovation and she did a good job with it. Go read it.
But here's what I want to add to my own take on it.
The framing isn't "AI is coming." The framing is: AI is already here, agencies are already using it, and carriers can't see any of it. That's the actual problem.
What agencies are doing vs. what carriers are seeing
Right now, most agencies are using AI for the basics. Email drafting, transcription, summarization, quoting support, administrative work. That's it. That's where AI is maximized today. And honestly, if you're not using it for those productivity gains, you're already behind.
But here's what I said in the article and I'll say it again here: there is no task in an insurance agency today that should happen without human oversight. Zero. AI right now is a copilot. Not an operator.
The problem is carriers don't know if agencies are treating it that way.
Delanea Davis, who works the carrier strategy side of this, put it plainly: carriers are asking whether submission quality improved, whether they're getting fewer calls, whether forms are being filled out more accurately. And in most cases, they're not seeing the movement yet. Either agencies aren't using AI the way they say they are, or they're using it in fragmented ways that aren't showing up in the metrics carriers care about.
That's not a technology gap. That's a trust gap.
The word the industry keeps missing: orchestration
Everybody wants to talk about governance. Transparency. Accountability. Fairness. Compliance.
Those things matter. But governance prevents bad things from happening. Orchestration is what allows good things to happen. And right now, the industry has very little of either.
I used this analogy in the article. Think about millions of cars suddenly appearing on the roads with no lanes, no signals, no navigation. The cars are the bots. Before we start arguing about speed limits, we need roads.
What Delanea described from the carrier side is essentially a call for a trusted middleware layer. Something like what PayPal or Zelle does in banking. They don't control everything before or after the transaction. They own one piece and authenticate the transfer. That's what insurance needs between agencies, AI systems, and carrier platforms.
Without that, carriers are stuck trying to govern workflows they literally cannot see.
Where accountability actually sits
This one's simple for me.
If something goes out under my agency's name, I own it. Full stop.
I have a license. I carry E&O risk. No quality agency owner I know is letting bots run end to end without licensed oversight. We have too much on the line.
But I also said something in the article that I think gets missed in these conversations: the bigger governance issue isn't rogue bots. It's agencies using personally identifiable information inside large language models without guardrails. That's where the real exposure is. Are people putting client data into systems they shouldn't? That's the question carriers and agency owners should both be asking right now.
What to do with this
Start with your workflows. Not with the technology.
Map your end-to-end processes first. Then ask honestly: where could AI support a step in this workflow without replacing the human judgment that step requires?
Winners route work. Stuck organizations pile work onto people.
That's not a new idea. It's just more urgent now.
Read the full article here: The Shadow AI Problem in Insurance. Jess did a good job with it.
And if you want the operational playbook for building an agency around routing, automation, and structure, that's what Future Ready is. No sponsors. No vendor pitches. Just the operating manual I wish someone had handed me five years ago.
