Property and casualty insurers are investing heavily in AI. BCG's AI Radar shows industry AI spending as a share of revenue will triple in 2026, yet returns are elusive; just 38% of P&C insurers are generating value at scale from AI in core workflows. The issue is not technical, however; it is strategic.
CEOs must push for a rethink. AI won’t deliver real value if dropped into legacy operating models designed for human-led execution. Instead, companies must redesign core insurance processes, such as underwriting and claims. This allows an insurer to transition to a fundamentally different execution model, with autonomous AI agents as the primary execution engine under human oversight.
It also opens the door to new, AI-native business models, such as an insurer-owned, AI-first distribution agency, unlocking premium growth.
Why Now
The economics of property and casualty insurance are under severe pressure. Pricing is soft, yet claim costs are rising due to inflation and increasing severity. Meanwhile, historical loss models and reinsurance capacity are under strain from climate risk. All this is happening as distribution models evolve rapidly, forcing insurers to improve speed, specialization, and digital execution.
This has led to a surge in AI experimentation; BCG data shows insurance leads many other sectors in AI pilots. However, its record on scaling to capture real value is one of the weakest. That’s because many insurers are focusing on tech-driven solutions without an effective strategy to capture cumulative value.
The value at stake is significant: our work with early adopters suggests that operating costs per dollar of premium could be reduced by 15% to 25%, which, in the US alone, equates to $35 billion to $60 billion in reduced operating expenses. Using AI can also increase premium per policy, as pursued by early adopters, and generate significant value; in the US, we see AI leaders winning an extra $8 billion to $20 billion.
However, the window of opportunity is already closing. Early movers are pulling ahead.
Stay ahead with BCG insights on artificial intelligence
How to Get There: Deploy, Reshape, Invent
Industry leaders are pursuing three linked plays:
Deploy. This is the application of AI to improve productivity in existing processes, such as summarizing claims submissions and providing action guidance in call centers. It is a short-term move that frees up capacity without changing core workflows.
Reshape. This is redesigning end-to-end workflows for the AI era. Orchestrated AI agents perform core tasks with humans engaged if exceptions trigger, providing oversight and judgment.
Reshaping helps retain more profit from each premium, generating meaningful improvements in areas such as leakage and recovery, claims handling costs, underwriting costs per policy, and policy servicing costs. But there are other benefits. AI-assisted intake and pre-fill can cut time-to-quote by 30% to 40%, a key competitive advantage.
Reinvent. This is where AI leaders are looking to use AI to change industry economics materially. An example would be an insurer-owned, AI-first distribution agency for standard- and mid-complexity risks, delivering higher margins and greater flexibility in pricing and product design.
Putting AI to Work
The 10-20-70 model provides a road map for resourcing. With this strategic approach, work on algorithms, although important, makes up just 10% of the effort.
Technology and data account for 20% of the total effort and constitute an essential foundation. It includes a shared ontology, a system of standardized meaning that is grounded in a business context, allowing AI models and agents to have an end-to-end interpretation of structured and unstructured data.
A full 70% of effort goes into creating an agent-first operating model, supported by talent and change management.
To redesign processes with an AI-first lens:
- Shift away from functional silos and create clear functional ownership.
- Build processes to be executed by AI agents, end-to-end.
- Refocus human expertise on validating AI outputs and resolving exceptions and high-impact outcomes.
- Enable real-time visibility and accountability to accelerate intervention and optimization.
The Path to AI Impact
To become an AI-first P&C insurer, CEOs and the C-suite must set the tone by:
- Defining an enterprise AI ambition embedded in enterprise strategy and linked to material economic outcomes.
- Driving alignment and outcomes with clear priorities, funding, and enterprise-level tradeoffs resolved.
- Upskilling and enabling the workforce by institutionalizing continuous skill evolution as agents assume a greater share of execution. Drive change management across the enterprise.
- Prioritizing the investments that deliver transformation rather than hundreds of disconnected proofs-of-concept.
For incumbent P&C insurers, the choice is becoming clear: lead the transformation or be forced to react to it. Those that move with genuine strategic ambition will define the industry’s next era. Those that don’t will find themselves competing on terms set by others.
