Insurance brokerages racing to adopt artificial intelligence may discover that years of loosely integrated acquisitions have left them without the operational foundation needed to use the technology effectively.
Curtis Barton (pictured), founder and CEO of California-based broker ALKEME, said brokerages need centralized data and standardized processes before AI can be applied consistently across an organization. That could present a particular challenge for acquisitive platforms that previously allowed agencies to retain separate systems and operating models.
“In order for AI to efficiently and effectively change the course of your business, you’ve got to have SOPs that you can run on rails on top of,” Barton told Insurance Business.
“Because if you don’t have a standardized SOP across your entire organization that everyone buys into, AI is going to be sporadically used in different ways.”
The warning comes as insurance brokerage M&A enters a more selective phase. OPTIS Partners recorded 292 US and Canadian agency transactions during the first half of 2026, down 15% from the same period in 2025 and the slowest start to a year since 2016.
PwC has similarly predicted that distribution deal volumes will moderate as buyers become more selective and integration capacity tightens. Investors are placing greater weight on organic growth, technology infrastructure, reporting capabilities and the execution of acquisition strategies, the firm said.
For Barton, those priorities represent a significant shift from the acquisition model that shaped much of the brokerage market’s recent consolidation. He said: “The requests from the private equity partners and the investors completely shifted from this high volatility, like buy as much as you can, to, ‘Hey, what’s your methodology? Where’s your organic growth? How integrated is your company? What is your command on your data?’”
Bringing acquired agencies onto common systems can be difficult after owners and employees have been promised operational independence. Barton said ALKEME adopted a fully integrated model from its launch, allowing the brokerage to focus technology investment on growth and servicing rather than bringing disparate operations together.
This is also shaping how ALKEME evaluates potential acquisitions. The brokerage has developed an M&A system that compares prospective agencies with historical transactions according to region, industry, valuation and post-acquisition performance.
“We have a lot of conversations and we don’t buy a lot,” Barton said. “It’s (about) doing the right deal, a meaningful deal, and making sure that it’s something that can feed our entire network.”
Barton doesn’t expect AI readiness to be well developed at most independent agencies. Bringing an agency into a larger platform may therefore require new controls around approved tools, access to data and the types of client information employees are permitted to enter.
“The first part of AI is getting your ducks in a row, getting your data in one spot, getting your SOPs established, and then running rails into what you want the result to look like,” he said. Brokerages can then identify repeatable administrative work that technology can support without removing employees from decisions that require judgment or client knowledge.
Even firms with significant technology budgets could struggle if employees do not incorporate the tools into their daily work. Barton compared the challenge with large software implementations, where the system’s value ultimately depends on adoption across the organization.
This makes the effect on account managers and producers particularly important. Barton said the opportunity lies in reducing routine tasks, so employees have more time to communicate with clients and examine gaps across their insurance programs.
“We want to make our people take the sludge off of their desk,” he said. “Get them focused in on doing highly paid jobs that they can actually get fulfilment out of.”
Barton expects AI adoption to increasingly influence productivity across the brokerage sector as firms move beyond experimentation and incorporate the technology into established workflows.
“AI’s not coming for your job, but the person that harnesses AI effectively will take your job,” he stressed.
