Insights from BrokerTech Connect on what it really takes to reimagine brokerage operations with artificial intelligence (AI) at the heart of the organization.

This summer at the inaugural BrokerTech Connect: Chicago event, AI was more than just a buzzword. It is reshaping how organizations think about work, talent, and value creation, and forcing leaders to confront whether their operating models are built for what comes next.

A wide-ranging discussion on AI-first organizations was moderated by Scott Weisman, co-founder and CEO of LaunchPad Lab. Joining him was Garrett Droege, former SVP - Director of Innovation + Digital Practice Leader at IMA Financial Group, who offered candid perspectives on what AI adoption really looks like inside complex insurance organizations.

Early Innings, Real Momentum

Drawing from more than a decade of building custom digital products, Weisman noted that many of today’s AI conversations are less about invention and more about execution. “There are these incredible tools on the delivery side and on the development side,” he said, pointing to how AI is dramatically accelerating product timelines. The real question, he added, is whether industries like insurance are positioned to translate that acceleration into meaningful outcomes. 

For many carriers and brokers, those early benefits are appearing first behind the scenes. Organizations are prioritizing internal use cases as they work to balance productivity gains with regulatory considerations and brand risk. 

Droege described how IMA has focused on using AI to organize and activate institutional knowledge. “We have a lot of documents that tell our people how to do what they do,” he said. “We built an LLM so that people could just chat with all these policies and procedures and it’s worked incredibly well.”

While proof-of-concept projects are proliferating, Weisman observed that many organizations he works with are still struggling to realize full market value from their AI investments. “They say, ‘We’re rolling out these proof-of-concepts and having success with smaller applications,’” he said, ‘“but we’re really not there yet when it comes to fully realizing the value of the solutions we’re paying for.’” His question to the group was direct: should organizations be pushing harder to move these tools into production, or does the pace of change need to remain measured?

From AI Tools to an AI-First Mindset

The panelists were careful to distinguish between deploying AI tools and embracing an AI-first operating philosophy. The latter, they argued, requires rethinking workflows from the ground up rather than layering technology onto existing processes.

“If we were building a brokerage today, it would probably not look anything like our firms,” Droege said. “We wouldn’t design it in the same way. We are trying to build that AI-first broker.” 

That aspiration reflects a broader shift in thinking. An AI-first strategy is not about automating individual tasks; it is about redesigning how information flows, how decisions are made, and where human expertise adds the most value.

Why End-to-End AI Still Has Limits

That tension between ambition and practicality was a recurring theme. The appeal of a fully reimagined, AI-first operating model is clear, but so are the complexities of the insurance industry. Weisman raised the question directly, “Does it need to be an end-to-end change with an AI-first mindset, or is there room for AI to be layered into legacy environments?”

“AI is not yet ready for end-to-end workflows,” Droege said. “It is ready for middle to middle. It still requires a human to prompt it, and you have to be really good at prompting.” He added that hallucinations remain a real concern, making expert review essential.

The implication is not to slow down, but to be intentional, matching use cases to both technological readiness and business impact.

Simplifying the Stack, Not Adding to It

Years of point-solution adoption have left many organizations with fragmented workflows that frustrate users and dilute returns. One of the promises of AI is the potential to reduce complexity rather than compound it. 

“I don’t want our people to have eight different tools in one workflow,” Droege said. “That’s a bad workflow.” 

Instead, he described a future where AI operates quietly behind the scenes. “I want one workbench that everybody lives in, and all these great tools that we’re using—no one knows the name of. They’re just powering the workflows.” 

In that model, AI becomes an orchestrator—not a destination—removing friction so brokers can focus on higher-value, client-facing work.

Adoption Starts from the Bottom Up

Weisman framed adoption not as a technology challenge, but as an organizational one. In his experience, companies often struggle less with capability than with prioritization. Many leaders, he noted, want to use AI more but are unsure where to start. “They don’t know what the first thing is to pick off,” he said. “What’s easy? What’s hard? How do you even begin to prioritize?” 

Droege advocated for empowering those closest to the work. “The people that are in the business are going to be the best sources to tell you what’s most beneficial,” he said. 

AI task forces and peer-led experimentation, he argued, can accelerate adoption more effectively than executive directives. “It’s not a CEO message saying, ‘We’re going to use AI,’” Droege said. “It’s seeing the person next to you take 30 percent of their day back.” 

Cultural resistance remains one of the biggest obstacles to AI adoption. When asked how he addresses it, Droege was candid. “I’m not going to replace anyone with a computer,” he said. “But I will expect every one of our colleagues to use AI every day. If you’re operating like it’s 1990, that’s just not a good place to be.” 

AI-first organizations will still rely on human judgment, creativity, and relationships, but those skills will be amplified by intelligence built into the workflow.

A Strategic Reset, Not a Technology Project

Throughout the session, Weisman returned to one consistent throughline: AI’s value is determined not by how advanced the technology is, but by how intentionally it is applied. Faster development cycles and cheaper experimentation, he cautioned, do not automatically translate into better outcomes. Without clarity on purpose and priorities, organizations risk building quickly without building what matters. 

The AI transformation will take time, but as the conversation made clear, the firms that treat AI as a strategic reset rather than a technology initiative will be best positioned to lead the next chapter of insurance innovation.