The insurance market is already shifting, and not in a subtle way. Larger insurers have stopped experimenting with AI. They are embedding it into the core of the business. You can see it in underwriting support tools, fraud detection, and document processing across the market. This is operational advantage, and it is starting to compound across the industry.
That creates a problem for SMEs.
While enterprise players strip friction from their workflows, many smaller brokers and insurers are still circling the idea of AI. There's interest and curiosity, but also hesitation. So the gap widens. On one side, firms that are improving how they work. On the other, firms still relying on manual coordination across email, spreadsheets and legacy systems.
This isn't a technology problem. Insurance has always been process-heavy, document-heavy and data-heavy, which is exactly why AI works so well in this sector. The challenge is that most firms don't know where to start in a way that delivers tangible value, so they default to doing nothing.
Meanwhile, the firms that are moving are embedding AI directly into underwriting, claims and operations. They are reducing time spent on repetitive work, improving consistency, and speeding up decision-making. AI is becoming part of the operating model, not an optional extra.
For SMEs, the pressure is more acute because the challenges are the same but the resources are not. Every broker or insurer still needs to process submissions quickly, manage complex information accurately, respond to clients, and stay on top of compliance. They just do it with leaner teams and less capacity.
That is where AI becomes practical rather than theoretical.
Used properly, it removes the work that does not need to be done by a human in the first place. In most insurance businesses, that tends to cluster around a familiar set of activities:
- Reviewing and triaging submissions
- Extracting data from emails and attachments
- Drafting routine client or insurer communications
- Summarising claims and case histories
- Pulling relevant information to support underwriting decisions
None of this is particularly futuristic. It is just work that currently consumes time and attention.
The impact is not just efficiency, it is capacity. Teams can handle more volume without increasing headcount, turnaround times improve, and the risk of error is reduced. More importantly, people spend less time coordinating work and more time on judgement, client relationships and commercial activity.
This is also where SMEs have an advantage, even if they don't realise it.
There is a persistent assumption that AI favours scale, but smaller organisations can often move faster. There are fewer layers of approval, less legacy complexity, and more room to test and adapt. Rather than reworking the whole business, SMEs can focus on a handful of workflows that slow them down and fix those first. That is often enough to close the gap with much larger competitors.
The difference between firms that succeed with AI and those that don't usually comes down to how they approach it. The wrong starting point is asking "how do we use AI?". The right starting point is much simpler: where are we losing time, adding cost, or creating unnecessary complexity?
Once that is clear, the application of AI becomes obvious. Good adoption tends to share a few common characteristics:
- Focused on real operational problems, not abstract use cases
- Fits into existing workflows rather than sitting alongside them
- Secure and aligned with regulatory expectations
- Built to support teams, not replace them
- Measured against outcomes like speed, accuracy and cost
If those things are not improving, the implementation is not working.
At North Stack, that is where we focus. Not on abstract strategy, but on what changes inside the business. We work with insurance firms to find where time is being lost and where processes are breaking down. Then we implement AI in a way that fits into underwriting, claims and operational workflows.
The objective is straightforward: reduce manual effort, improve speed, and create capacity without increasing cost. Done properly, that makes the business easier to run and more resilient as volumes grow.
The direction of travel across the market is already clear. Larger insurers are seeing the benefits and building AI into how they operate. That creates competitive pressure that will increasingly be felt by SMEs.
The question is no longer whether AI will matter. It is whether firms move early enough to benefit from it.
Because the firms that act now will be able to operate more efficiently, respond faster, and scale without the same level of friction. Those that wait will still be dealing with the same operational constraints, just in a market that has moved on.