Dearborn Labs launched to embed AI across insurance operations
March 13, 2026
Clearcover, the Chicago-based AI-native auto carrier, has launched Dearborn Labs - a forward-deployed AI practice that builds and runs production AI systems inside property and casualty insurers and MGAs. The goal: move past isolated pilots and wire AI directly into underwriting, claims, and distribution so it drives measurable results, not slideware.
What Dearborn Labs actually does
Dearborn Labs partners with carrier and MGA teams to design, build, and deploy customised AI systems inside live environments. It's not a static platform. It's an operating layer that adapts models, data, and workflows to each insurer's business model.
Kyle Nakatsuji, Founder, Dearborn Labs, said, "Most carriers have already invested in AI. The problem isn't the tools. It's that the landscape changes faster than any single solution can keep up with, and nobody's connecting the data and context across operations to make those tools compound. That's not a software problem. It's an operating problem."
Why this matters for insurance ops
Instead of standing up yet another system, the practice integrates existing data and processes across departments. The intent is to make claims data inform underwriting, and underwriting context shape distribution - creating compounding effects across the value chain.
As Nakatsuji put it, "We don't hand over a strategy deck. We deploy into your operation and ship production systems in weeks. Your claims data should make your underwriting smarter. Your underwriting context should shape your distribution. We build the infrastructure that makes that happen."
Early performance signals from Clearcover
Clearcover reports material lift from its internal AI stack across multiple states. More than 90% of claims intake runs through AI agents, 93% of policies are bound digitally, and claims handling operates at roughly three times the efficiency of traditional carriers.
Where carriers can put this to work
- Underwriting: Prefill, risk signals from claims and telematics, triage for referral vs. straight-through, pricing context back into marketing.
- Claims: AI-guided FNOL, coverage validation, liability assessment support, fraud signals, subro identification, and faster settlement routing.
- Distribution: Lead scoring, next-best-action, appetite feedback loops from underwriting to agents and digital funnels.
- Operations and data: Cross-system orchestration, model monitoring and retraining, and feedback loops that cut cycle time and expense.
Business outcomes on the table
- Lower loss ratio through better risk selection and earlier fraud detection.
- Improved underwriting precision with richer, real-time context.
- Reduced LAE and operating expense via automation and straighter-through flows.
- Shorter cycle times and a cleaner customer experience.
How engagement works
- Weeks, not quarters: Deploy into production with measurable KPIs and guardrails.
- Use what you have: Integrate current systems, models, and data rather than replacing them.
- Prove and scale: Start with a contained workflow (e.g., FNOL or bind flow), validate gains, then expand.
Dearborn Labs is accepting a limited number of carrier and MGA engagements for Q2 2026.
For leaders planning the next step
If you're evaluating how to embed AI into underwriting, claims, and distribution without adding more shelfware, start with a single workflow tied to a clear KPI (cycle time, digital bind rate, LAE) and build the data feedback loop from day one. This is the compounding effect Dearborn Labs is optimising for.
For broader industry context and practical playbooks, see AI for Insurance. Learn more about Clearcover's approach to digital insurance at clearcover.com.
Your membership also unlocks:
