Ten percent.
That's how many P&C insurers are actually scaling AI right now. Not piloting it. Not exploring it. Scaling it. And according to Capgemini's 2026 World P&C Insurance Report, that 10% is pulling away from the field at a rate that's starting to look permanent. They're seeing 21% higher revenue growth and 5.1% greater share price increases compared to their peers.
The other 90% are still in the exploration or proof-of-concept stage. Forty-two percent aren't measuring AI outcomes at all.
This week, the tools got better. The gap got wider. And the window for independent agency owners to get on the right side of it got a little bit smaller.
———
THIS WEEK'S INSIGHT: The Governance Gap Is the Real Story
Everyone wants to talk about AI capability. The Capgemini data points somewhere else entirely.
The agencies and carriers pulling ahead aren't winning because they found a better tool. They're winning because AI is embedded in how they operate, not bolted on as an experiment. The trailblazers are nearly four times more likely to invest in change management beyond basic training, nearly three times more likely to have explainable AI infrastructure driving enterprise-wide confidence, and nearly twice as likely to have AI responsibilities embedded directly into job descriptions.
That last one is the tell. When AI is someone's job, it gets measured. When it gets measured, it improves. When it improves, it compounds.
Meanwhile, 44% of insurance executives say governance or compliance challenges have contributed to AI projects failing or underperforming. Only 24% are confident they could pass an independent AI governance review in 90 days.
The agencies stuck in the 90% don't have a technology problem. They have an ownership problem. Nobody owns the AI, so nobody improves it, so nothing compounds.
That's the real gap. And it's not closing on its own.
———
AI NEWS THAT MATTERS
→ Anthropic doubled Claude Code rate limits across all paid tiers
Anthropic removed peak-hour throttling and doubled throughput at no added cost. For agencies building internal AI tools on Claude, this removes one of the last friction points standing between a working prototype and a production deployment. If you've been holding off on building because of capacity concerns, that excuse is gone.
→ OpenAI pushed GPT-5.5 Instant as the new ChatGPT default
The update ships stronger cross-session memory to all web users. For insurance agency owners, this means a prospect-facing intake tool built on ChatGPT can now remember prior conversations, policy context, and uploaded files without the client re-explaining their situation each time. Persistent client context without custom infrastructure is newly viable today. That's a meaningful shift for any agency exploring AI-assisted intake or triage.
→ The 10% are pulling away
Consumer support for AI in insurance nearly doubled year over year, from 20% in 2025 to 39% in 2026, while resistance eased from 44% to 36%. The public is warming up. The window for early movers is open. The question is whether your agency is moving.
———
EARLY MOVER ADVANTAGE
The resistance is easing faster than anyone predicted. Consumer comfort with AI in insurance nearly doubled in a single year. That means the agencies that get AI embedded now aren't just building operational leverage. They're building brand trust with a customer base that's increasingly ready to accept it.
The early mover advantage in AI isn't just about efficiency. It's about positioning. When your agency is already running on AI infrastructure and your competitor is still exploring chatbot pilots, you're not competing on the same playing field anymore.
The agencies that move now capture the adoption window. The ones that wait inherit the compliance requirements without the head start.
———
TOOL SPOTLIGHT: Signal vs. Kendo
Two tools. Two different problems. Two different eras.
Kendo gives your team one-on-one training against human-like AI that emulates real-world sales calls, scores and analyzes them, and helps train smarter so you can keep hitting sales targets. It's a strong tool. Insurance teams using Kendo's daily AI roleplays have reported 20-25% ROI, with reps practicing against realistic personas that mirror real objections before live calls. Pricing starts at $55/seat/month. If your producers need to sharpen their skills, Kendo is built for that.
Signal is a different problem entirely.
Signal doesn't train your humans. It scores your AI agents. The SAUCE scoring system evaluates every agent output across four dimensions: Logic Soundness, Source Accuracy, Problem Solving, and Adaptability. The fifth dimension, Intelligence Density, measures quality per token cost against a benchmark. The Agent Arena runs three live leaderboards reset weekly — Triage, Outreach, and Follow-up — so you always know which agents are performing and which need work.
Kendo answers the question: how good are my reps?
Signal answers the question: how good is my AI workforce?
If you're still training humans to do what AI should be doing, Kendo is the right tool for right now. Signal is what comes after. It's built for the agencies that have already made the shift and need infrastructure to measure and compound it.
Signal is live. Early access is open at closemodeai.com.
———
THE BOTTOM LINE
The 10% aren't waiting for the industry to catch up. They're building infrastructure that makes catching up harder for everyone else.
The free Agency AI Audit at closemodeai.com takes eight minutes and shows you exactly where your agency stands. Where the gaps are, what can be automated now, and what a CloseMode AI build looks like for your specific operation.
No pitch. Just signal.