Pennsylvania Lumbermens Mutual Insurance Company (PLM) is resisting the industry-wide rush to bolt AI onto every process, with COO Lindsey DiGangi warning carriers to avoid what she calls the "FOMO" trap. The specialty mutual, which serves the wood and building materials industries, is instead focusing on targeted modernization in areas it deems most critical - a measured approach that contrasts with the broader scramble across commercial insurance.
"We're not going to throw it at everything, but we can find the low-hanging fruit and start modernizing in the areas that we feel are most critical," DiGangi said. "There's always a little bit of FOMO, and it's more or less just that opportunity is knocking, and how do we decide where to start?"
Data connectivity as the first win
PLM has long operated on a buy-versus-build model for technology - a practical choice for a carrier its size, DiGangi said, but one that scattered data across disparate systems. The early priority has been pulling that information together. "Early wins have been about connecting all that data into one place and starting to redefine the infrastructure of data at PLM to make it more accessible," she told Insurance Business.
Access alone is not enough. DiGangi said the company is investing in education across the organization, helping employees understand what questions to ask of the data now available to them. "You kind of have this 'don't know what you don't know,'" she said. "It's all there available to us, but how do we figure out what we should be seeing and what we could be learning from what's here?" Many carriers sitting on decades of information face the same bottleneck - the gap between having data and extracting practical insight from it. This challenge is central to broader discussions around AI for Insurance, where data readiness often determines whether AI tools deliver value or disappointment.
Governance and the room to experiment
In a regulated market, governance is non-negotiable. DiGangi said insurers need space to test new approaches but must operate with appropriate controls. The balancing act - experimentation within guardrails - is one many mutuals are working through as they update systems without abandoning the underwriting discipline that defines the model.
PLM's insureds are also changing. Companies that once operated in narrow segments of the wood and building materials supply chain are now expanding and diversifying. "For us, when we're looking at a risk, we're recognizing that these companies are much more complex than they once were," DiGangi said. "Businesses have learned that what consumers want is changing, so they have to find new ways to bring value to their customers. They are kind of reinventing themselves up and down the supply chain." Insurance products must keep pace, she added, but risk management is just as important - the industry has shifted from a reactive posture to a proactive, enterprise-wide view of risk.
Talent and the long view
Talent remains a long-term concern. DiGangi said insurance companies need to rethink the roles and skills they require as the industry changes. The traditional workforce model will need to shift alongside the risks carriers are being asked to cover. For AI for Executives & Strategy, this means decisions about technology adoption are inseparable from decisions about who will use it and how.
DiGangi sees mutual insurers as well positioned for this moment. Their stakeholder-driven structure and long-term focus allow them to respond to market changes without chasing every trend. "We're a very stable company," she said. "We stay very committed to underwriting discipline and to risk management with our insureds. But we're not afraid of change and evolution as well."
Why this matters for insurance professionals
PLM's approach surfaces a practical truth: AI adoption does not require a blanket rollout to be effective. For insurance professionals - particularly those at mid-size and specialty carriers - the takeaway is that connecting existing data and educating staff on how to use it can deliver immediate value without the cost and risk of enterprise-wide AI deployment. The harder work is not buying the tools. It is building the data infrastructure and organizational know-how that make those tools worth buying in the first place.
Your membership also unlocks:
