Datos ILTF Frames the Intelligent Insurer Operating Model: Testing the Framework Against What Carriers Have Actually Built

Datos Insights introduced its Intelligent Insurer Operating Model at the April 2026 Insurance Leaders Technology Forum, built on survey data from 36 senior carrier technology leaders. It prescribes five pillars for restructuring a carrier around AI rather than bolting AI onto existing workflows.

The prescription is only useful against what the most advanced carriers have actually built. Travelers, Chubb and AIG have each implemented a different subset, and the pillar none of them is being scored against is the one that matters most.

Key Takeaways

• Roughly 60% of carriers have AI in production against 7% to 10% at meaningful scale, a 50-point gap that is the space the framework claims to occupy.
• Only 8% of the forum's respondents believe they lead peers on AI capability, while 70% expect at least moderate competitive advantage within three years.
• The three implementations diverge on which pillars they satisfy. Travelers leads on scale and speed, Chubb on work redesign and business leadership, AIG on orchestration.
• Governance is not one of the five pillars. AIG's agents now run up to 30 hours autonomously, against Grant Thornton's finding that 24% of carriers could pass an independent governance review in 90 days.
• 70% of carriers spend under $500,000 a year on AI, so the modular architecture the framework implies is being assembled from vendor components rather than built.

What the Framework Prescribes

The model's core proposition is that linear, siloed workflows should give way to coordinated human and AI execution across the value chain, and that the operating model has to be designed before the pilots rather than after them.

The five pillars run: scale without headcount growth, with agentic systems absorbing volume while people move up the value chain; intentional AI rather than experimentation, with operating model design preceding deployment; reimagine work rather than digitize it, on the argument that automating a broken process produces a faster broken process; business-led and technology-enabled transformation; and speed as the competitive variable rather than strategic positioning.

The survey behind it explains the framing. Only 8% of respondents believe they currently lead peers in AI capability, while 70% expect at least moderate competitive advantage within three years. That is a large population committed to an outcome it has not yet demonstrated, and the framework's pitch is an organizing principle for converting scattered production deployments into something coordinated.

The scale numbers set the target. Across surveys, roughly 60% of carriers have AI in production while 7% to 10% have reached meaningful scale, a gap of about 50 points.

Three Carriers, Three Different Subsets

Framework Pillar	Travelers	Chubb	AIG
1. Scale Without Headcount	Strong: 50%+ claims STP, call centers halved	Strong: 85% automation target, 20% headcount reduction	Strong: 4x submission processing with AI
2. Intentional AI	Strong: $1.5B tech budget, 8-year ramp	Strong: structured 3-4 year plan	Moderate: rapid expansion across 8 LOBs
3. Reimagine Work	Moderate: infrastructure-first, some workflow redesign	Strong: global claims restructure under single authority	Strong: multi-agent orchestration redesigns workflows
4. Business-Led	Partial: technology-driven with business outcomes	Strong: CEO-led, firsthand evaluation	Moderate: technology architecture leads
5. Speed	Strong: real-time quoting, STP	Secondary: discipline over speed	Strong: 55% time-to-quote reduction

Travelers has built infrastructure first. It commits $1.5 billion annually to technology with strategic AI spend more than doubling over eight years, has equipped nearly 10,000 engineers and data scientists with Claude assistants and more than 20,000 employees with AI tools generally, and runs over 50% of claims as straight-through-processing eligible with two-thirds of eligible customers taking it. Claims call centre population is down by a third and locations have gone from four to two.

That is strong on pillars one and five and partial on pillar four. Travelers accumulated 65 billion clean data points over decades and built the AI strategy on top of that asset, which is technology-driven with business outcomes rather than the business-composed sequencing the framework prescribes.

Chubb inverts it. The target is 85% automation of major underwriting and claims processes over three to four years, with workforce reductions of up to 20%, roughly 8,600 of 43,000 employees, and run-rate savings of 1.5 combined ratio points; about 85% of global gross written premium already runs through digital or significantly digitally enabled channels. The appointment of Kevin Rampe as the first Global Claims Officer across 54 countries is the structural move pillar three describes, since coordinated human-AI claims workflow is not deliverable without a single point of authority over 54 regulatory environments.

Where Chubb departs is speed. Greenberg's emphasis is on firsthand executive evaluation rather than time-to-market, which is a rational trade at an 84% combined ratio: the downside of disruption scales with what there is to protect.

AIG maps to the framework most completely and is the most architecturally aggressive. AIG Assist runs across eight lines of business with purpose-built agents for submission ingestion, risk evaluation and pricing benchmarking, coordinated by an orchestration layer that controls activation, information sharing and human oversight triggers. Lexington reported a 30% increase in quoted submissions, a 55% reduction in time to quote and approximately 40% more binding of submitted business, with AIG processing four times more submissions with AI assistance.

The Pillar That Is Missing

Governance is not one of the five, and the gap shows most clearly in the implementation that scores best.

Agent autonomy at AIG has extended from under one hour to up to 30 hours of continuous operation, which Peter Zaffino says evolved faster than expected over nine months. The framework treats organizational change management as a parallel implementation requirement rather than a pillar, so a carrier can satisfy all five while running unsupervised agents for a working day and a half. Grant Thornton's finding that only 24% of carriers could pass an independent AI governance review within 90 days indicates the omission is not theoretical.

The vendor layer carries the same problem in a different form. Kurt Diederich of Finys argued in Carrier Management that carriers should treat AI as a modular capability inside a plug-and-play operating model, replaceable with minimal disruption, drawing the parallel to the early 2000s internet proliferation where only a small share of vendors proved durable.

The framework's second pillar points the same way, but modularity is an organizational capability rather than an architectural preference. It requires defined interfaces between components, standardized data formats and governance spanning multiple vendor relationships. With 70% of forum respondents spending under $500,000 a year on AI, most are buying vendor solutions rather than building the abstraction layer that would make them swappable, and OpenAI already sits in 90% of carrier stacks.

So the operating model the framework describes and the operating model carriers are assembling differ in their failure mode. The prescribed version fails when the workflow design is wrong. The assembled version fails when a vendor is acquired, pivots, or breaks, and the pillar that would have caught it is not in the model.