Gennaro Brooks-Church on the Line He Won't Let AI Cross

Gennaro Brooks-Church, Founder and CEO of Cazimir and Brooks Energy, spends as much time talking about what AI shouldn't do in insurance as he does about what it can. In an industry racing to adopt AI faster than it can fully understand it, Gennaro Brooks-Church has drawn one line he refuses to cross, no matter how capable his systems become.

He has said it plainly more than once: AI should never be the one making the final call on complex underwriting decisions. It should support that decision, not replace the person making it. That single idea shapes nearly every product decision at Cazimir and Brooks Energy, even when a more automated approach might look more impressive on paper. This isn't a minor detail buried in documentation. It comes up unprompted, in interviews, client conversations, and how he describes his own work.

Underwriting depends on judgment that's hard to fully explain, let alone automate. An experienced underwriter isn't just following a checklist, they're weighing dozens of small signals at once, some they might not even be able to articulate if asked. Gennaro Brooks-Church has built his approach around respecting that reality instead of engineering around it. He's talked about the real opportunity with AI not being speed for its own sake, but making sure knowledge that took decades to build doesn't disappear the moment someone leaves the industry. A system that took over final decisions entirely wouldn't just risk getting things wrong, it would risk erasing the expertise it was supposed to protect.

Rather than pushing his systems toward full decision-making authority, Gennaro Brooks-Church designs them to handle the parts of underwriting that don't require judgment at all, document classification, data extraction, flagging inconsistencies across a submission. The moment a decision requires weighing risk or context, the system hands responsibility back to a person. This isn't a limitation he's working to remove over time, it's intentional, and one of the harder disciplines to maintain given industry pressure to automate more and market it as innovation.

It would be easier, in some ways, to loosen this line. Competitors often promote how much decision-making their tools can handle, and there's real commercial incentive to follow that trend. But Gennaro Brooks-Church has resisted it. Underwriters won't rely on a system long-term if they suspect it's quietly taking over decisions they should be making themselves. He measures success by whether teams keep using a tool on ordinary days, not by how impressive a demo looks.

Holding this line isn't just stated intention. Every output his systems produce traces back to its original source document, so underwriters can verify information themselves rather than accepting a conclusion outright. This traceability reinforces the boundary between support and replacement, keeping the human reviewer in the decision-making seat.

For Gennaro Brooks-Church, this line is a deliberate definition of progress, not a temporary limitation. Success isn't measured by how much human involvement a system eliminates, but by how much institutional knowledge the industry keeps, and how much trust a tool earns from the people using it daily. In an industry crowded with founders promising full automation, he has built a reputation around restraint as much as innovation, making sure expertise never has to disappear in the first place.