P&C Insurers' 72/28 AI Spend Split Exposes an Architecture Mismatch

Capgemini's 2026 World Property and Casualty Insurance Report surfaces a spending pattern that rewrites the AI narrative for the sector. Insurers direct 72% of AI budgets to technology and just 28% to change management. The split itself is not the headline. The headline is that 42% of insurers track no AI metrics at all, which means they have no systematic way to validate what works, no playbook to scale it, and no mechanism to stop what isn't working.

The simple read says insurers are finally embracing artificial intelligence. The better market read says the industry is funding capability without building the organizational wiring to convert that capability into performance. Capgemini calls this an architecture mismatch – a structural gap that runs deeper than the technology stack and that no amount of incremental AI investment will close on its own.

For investors tracking P&C carriers, the mismatch is a filter. Companies that treat AI as a program to be managed will see diminishing returns from rising budgets. The few that treat it as a core operating capability – and align strategy, technology, and organizational adoption in tandem – are already pulling away. The report identifies the top 10% as intelligence trailblazers. Over three years, that group delivered 21% higher revenue growth and 51% greater share price increases compared with the rest of the industry.

The 72/28 Spending Split That Defines the AI Gap

The allocation imbalance is the first signal that the constraint is organizational, not technological. Technology creates capability. Change management, however, determines whether that capability becomes performance. When nearly three-quarters of the budget flows to tools and platforms while barely a quarter goes to adoption, workflow redesign, and talent realignment, the result is predictable: pilots multiply, yet day-to-day work barely shifts.

Capgemini's data confirms the stall. 47% of employees report no meaningful change in their day-to-day work after 18 months of using AI. That is not a deployment failure. It is a design flaw. Insurers built the technical rails, then left the human and process architecture untouched. The spending split is the mechanism that locks in that flaw.

Key insight: The 72/28 split is not a budgeting oversight. It is the financial expression of an assumption that AI is a technology problem. Until that assumption is reversed, more spending will produce more pilots, not more performance.

Why the Split Persists

The split made sense when AI was unproven and the organizational implications were unclear. Legacy systems were the right investment at the time. Prioritizing technology over change management was a rational choice in a different context. The problem is that the context has changed, and most organizations have not systematically asked whether the investments already made still pay back on the original terms.

Without that question, the architecture underneath the pilots stays frozen. New tools are deployed on top of old decision structures, old accountability lines, and old workflows built for human execution. The spending split endures because it is the accumulated consequence of individually rational choices, not evidence of poor judgment.

Why 42% of Insurers Can't Measure AI Success

The absence of metrics is the second signal that the mismatch is structural. When 42% of insurers track no AI metrics, they have no feedback loop. They cannot distinguish a pilot that is failing quietly from one that is ready to scale. They cannot redirect capital from underperforming initiatives to high-return ones. They cannot build an institutional playbook because they have no record of what worked.

This measurement vacuum is not a data problem. It is an accountability problem. Who defines success? Who is responsible for outcomes beyond deployment? How is progress measured when decisions were designed for a world where underwriting, claims, and distribution judgments were quintessentially human? Most organizations have not answered those questions. The 42% figure is the consequence.

Risk to watch: Insurers that continue to deploy AI without metrics will eventually face a credibility gap with boards and investors. When the spending cycle turns, the absence of a validated ROI story will make AI budgets an easy target.

The Three-Dimensional Architecture Mismatch

Capgemini's report breaks the mismatch into three entangled dimensions. Addressing one while leaving the others untouched limits progress rather than unlocking it. The conventional sequence – fix strategy, then technology, then organization – fails precisely because the dimensions are interdependent.

• Strategy and talent: Only 35% of top 20 global P&C insurers have explicitly linked AI strategy to business outcomes beyond efficiency. That narrow framing directs investment toward quick wins rather than the capabilities AI needs to compound over time.
• Technical constraints: Legacy architectures fragment data across underwriting, claims, and distribution, making it harder for AI to reason across context-rich, unstructured information. The barrier is less about the AI itself and more about an environment not designed for it.
• Organizational ownership: 55% of insurers cite unclear ownership of AI initiatives as a key constraint. Without clear accountability, programs stay dependent on individual champions rather than building institutional capability.

Strategy and Talent: Only 35% Link AI to Business Outcomes