AI is accelerating development, reshaping customer service, and uncovering opportunities hidden inside brokerage data. But realizing that value requires specificity, strong data governance, and a clear understanding of where technology ends and human accountability begins.
"Innovation." "Transformation." "Disruption." "Automation." "AI."
Insurance leaders hear these words so often that they've begun to lose meaning.
At a recent BrokerTech Connect panel, moderator Gene Chao, CEO of FormativGroup, compared these industry favorites to a household curse-word jar. The words themselves are not bad, he noted. The problem is that they often become meaningless when nobody defines what they are expected to accomplish.
That observation captures one of the most important realities facing brokerages today: success with AI depends less on having a bigger AI strategy than on being more specific about the business problem, the data, the workflow, the economics, and the role of human accountability. The brokerages that create meaningful value from AI won't necessarily be the ones that adopt the most technology. They'll be the ones that connect technology to clear business outcomes.
A New Way of Thinking About Workflows
For decades, organizations operated under a familiar model: people followed processes that were supported by technology. According to Chao, that dynamic is changing. Increasingly, technology is running more of the process while people provide governance, judgment, and intervention when expertise is required.
That doesn't make people less important. It changes where their value is concentrated. That shift also raises the standard for oversight: reviewing an AI-driven outcome only adds value when the reviewer has the context, time, and authority to question it.
Chao illustrated the shift with a comparison many business leaders can relate to. Traditional workflows looked a lot like MapQuest: predefined routes with limited options. Modern workflows increasingly resemble Waze, constantly adapting to new information, changing conditions, and real-time signals. As AI becomes more deeply embedded into brokerage operations, the question is no longer whether technology can support a workflow. The question is how people and technology should work together within that workflow.
Start With the Outcome, Not the Tool
One of the panel's strongest messages was that AI is not a single thing.
It includes user interfaces, machine learning models, reasoning engines, automation capabilities, and the underlying knowledge created from data. Yet many organizations still begin their AI journey by evaluating products before clearly defining the problem they're trying to solve.
A more productive approach starts with basic questions:
- What business outcome are we trying to achieve?
- What workflow do we want to improve?
- What concept is a proof of concept actually intended to prove?
- What data will be required?
- What action should occur once an insight is generated?
- What will it cost, and how will we know it delivered a return?
As Stephen Siegel, Senior Advisor, Insurance at FormativGroup, put it, "If you're doing a proof of concept, know what concept you're trying to prove."
That level of specificity helps organizations move beyond experimentation and toward measurable value.
Where AI Can Create Real Brokerage Value
While AI discussions often focus on the technology itself, the panel emphasized practical business outcomes across three areas: growth, retention, and operations.
- Growth: Uncovering Opportunities Hidden in Data
The modern brokerage sits on an enormous amount of information, especially as industry consolidation creates larger organizations with multiple offices, specialties, and lines of business. According to Siegel, AI can help brokerages identify cross-selling opportunities that may otherwise remain invisible. Acquired capabilities, product lines, and specialized expertise often exist somewhere within the organization, but many firms struggle to connect those opportunities with the right clients.
AI can also identify external and internal changes that signal evolving client needs. Interest rates rise. Property values change. Businesses acquire new locations. Families experience marriages, births, home purchases, or retirement transitions. Each event may create an opportunity for a coverage review or new product recommendation. Rather than waiting until renewal, brokerages can proactively engage clients when those needs emerge.
- Retention: Thinking Beyond Renewal
Many organizations view retention as a renewal activity. The panel challenged that assumption. Evan Groot, Global Go To Market Director at Salesforce, argued that retention is actually "the entire value chain," encompassing everything from the initial risk selection process to onboarding, service interactions, and ongoing client engagement.
AI creates an opportunity to identify warning signs earlier in the customer lifecycle. Service inquiries, email interactions, chat conversations, and call-center activity can all contain signals that a customer may be frustrated, considering alternatives, or preparing for a major change. By analyzing those interactions collectively, AI can help teams identify hidden intent and intervene before a relationship deteriorates.
The goal is not simply to respond faster. It is to understand what a customer may need before they explicitly ask for it.
- Operations: Automating the Routine, Elevating the Human
Operational efficiency remains one of AI's most immediate applications. Today's models can process information from documents, PDFs, spreadsheets, voice interactions, emails, and images at a scale that would have been difficult to imagine even a few years ago.
The impact extends beyond operational tasks. Groot shared an example in which a brokerage technology project that originally required 49 weeks to complete was recreated as an initial prototype in less than 30 minutes using frontier AI models. A more functional platform-based version was developed in just hours. While those outputs still required refinement and governance, they demonstrated how dramatically development cycles are changing. As routine work becomes more automated, employees can spend more time on judgment-intensive work, complex exceptions, and client relationships.
Trusted Data Is the Foundation
If there was one cautionary theme that surfaced repeatedly throughout the conversation, it was the importance of trusted data. An impressive answer is not necessarily a reliable answer. Organizations need to understand where data originates, what systems an AI model can access, how information moves through a workflow, and whether outputs can be traced back to their source.
The panelists emphasized the importance of auditability, governance, and domain-specific context. That distinction becomes especially important in regulated industries such as insurance, where recommendations, coverage decisions, and client interactions carry legal and compliance implications. Siegel offered a memorable analogy, describing large language models as "somebody you meet at a party who sounds like they know what they're talking about until they're talking about something you know about."
In other words: AI can be extraordinarily powerful, but confidence should never be mistaken for accuracy.
Human-First Versus AI-First Is the Wrong Debate
Debates around AI usage often become polarized around a false choice: should humans remain at the center, or should AI take the lead?
The panel suggested a more useful framework. Some interactions, such as document requests or basic service transactions, may eventually be handled almost entirely by AI. Others, including coverage recommendations, regulated decisions, beneficiary changes, or relationship-driven advisory conversations, require meaningful human oversight. The appropriate level of human involvement should reflect the risk and consequence of the task. Rather than replacing relationships, AI should strengthen them.
By surfacing client insights, identifying emerging needs, and helping brokers anticipate questions before they arise, AI can make advisors more informed, more proactive, and ultimately more valuable.
Two Questions Every Brokerage Should Ask
As the discussion concluded, Chao offered two simple calls to action:
- Be specific: Don't start with AI. Start with the use case, the workflow, the business outcome, and the boundaries around success.
- Know and manage your data: Understand where it comes from, where it goes, who can access it, and how it supports decisions that can be explained and defended.
The brokerages that gain the greatest advantage from AI won't necessarily be those with the most pilots, the most vendors, or even the most advanced technology. They'll be the organizations that deliberately connect trusted data, governed workflows, responsible action, and human expertise to better outcomes for their clients.
And that starts by moving beyond buzzwords.
Thursday, November 5, 2026