P&C carrier AI spending tripled as a share of revenue in 2026, per BCG's AI Radar, and only 38% of carriers are generating value at scale from AI in core workflows. BCG's own framework assigns 70% of transformation outcomes to people and process rather than algorithms, which locates the shortfall: the budget is going where the returns are not.
Key Takeaways
- Spending tripled as a share of revenue while 38% reach scale. The 62-point gap sits between tooling investment and the workflow redesign that converts it.
- BCG's 10-20-70 split puts 70% of success in people and process, the component that is hardest to book as AI spend and easiest to underfund.
- Straight-through processing runs at 14% of carriers against 36% planning it, and the 20% cost reduction target depends on STP at 60 to 80%.
- Analytics leaders posted combined ratios six percentage points lower and premium growth three points higher than slower adopters across 2022 to 2024.
- Forty-two percent cite data quality as a major obstacle, only 20% have a defined analytics strategy, and just 12% train staff regularly. None of those is a technology problem.
The Spend and the Scale Rate
BCG's March 2026 report on the AI-first P&C insurer sets the prize at roughly 20% cost reductions and 3% to 5% GWP growth through pricing precision and productivity, worth $35 to $60 billion of US P&C operating cost reduction. Tripling AI spend as a share of revenue in a year is a genuine reallocation of operating budget rather than incremental pilot funding, so the investment intent is not in doubt.
Where that money goes is the issue. BCG's 10-20-70 framework splits transformation success into algorithms at 10%, technology infrastructure at 20%, and people and process at 70%. Vendor contracts, model licenses, cloud compute, and data engineering absorb most incremental dollars, which is the 30%. Workflow redesign, change management, retraining, and governance sit in operating budgets where they are harder to label as AI investment, and they are the rate-limiting constraint.
P&C amplifies the mismatch. Underwriting and claims are not high-volume transactional processing in the way consumer banking is; they carry judgment calls by licensed professionals, incomplete loss information, adverse selection, and multi-year development. A tool inserted into a workflow built around sequential human decisions speeds one step while the handoffs around it stay manual, so document processing gets faster and total cycle time does not move.
Where the 38% Actually Sits
WTW's 2026 Advanced Analytics and AI Survey, covering 59 US and Canadian P&C insurers, maps the same finding at function level. The adoption curve is tiered, and the tiers explain the headline.
Advanced rating and pricing models are near-universal at roughly 80% of carriers with another 11% planning adoption. That is mature actuarial practice rather than recent AI investment, and it is the baseline nearly everyone has passed. Claims analytics is the opposite: 33% use advanced analytics for fraud detection and 29% for claims severity assessment, both projected to reach 65 to 70% within two years. Straight-through processing, the clearest marker of end-to-end integration, sits at 14% with 36% planning it, and AI augmentation of underwriting decisions at 16% with 60% prioritizing it by 2028.
Those numbers land almost exactly on BCG's 38%. The carriers at scale are the ones past pricing model maturity and into claims analytics, underwriting augmentation, and straight-through processing. The 62% are clustered in the middle: beyond basic pricing analytics, short of the integrated workflows that move cycle time and combined ratio.
| Metric | Analytics Leaders | Industry Average | Source |
|---|---|---|---|
| Combined ratio advantage | 6 pts lower | Baseline | WTW, March 2026 |
| Premium growth advantage | 3 pts higher | Baseline | WTW, March 2026 |
| Time-to-quote reduction | 30-40% | Minimal | BCG, 2026 |
| Underwriter active handle time | 30-40% reduction | Minimal | BCG, 2026 |
| Claims STP rate (target for leaders) | 60-80% (personal/std. commercial) | 14% current industry | WTW / BCG, 2026 |
The gap shows up in results rather than only in capability. Carriers with sophisticated analytics posted combined ratios six percentage points lower and premium growth three points higher than slower adopters across 2022 to 2024. Sustained across an underwriting cycle, six points compounds: the leader writes more business at the same margin, or holds price where peers are underpricing.
BCG's workflow data shows where those points come from. AI-assisted submission intake and pre-fill cuts time-to-quote by 30 to 40% for standard risks, and underwriter active handling time per account falls 30 to 40% on risks routed through AI-assisted triage. AIG reported processing time for complex commercial submissions dropping from three weeks to three hours. The expense arithmetic only changes structurally at 60 to 80% straight-through processing for personal and standard commercial lines, where fixed underwriting infrastructure carries substantially more premium without proportional staffing. That is the mechanism behind the 20% target, and it does not operate at 14%.
The Barriers Are Not Technological
Three constraints recur among carriers that have not scaled, and none of them is fixed by buying a better model. Forty-two percent cite data quality and accessibility as a major obstacle, only 20% have a well-defined analytics strategy, and just 12% regularly train staff on analytics.
Data quality is the load-bearing one. Models trained on fragmented, inconsistent history produce outputs that underwriters and claims handlers correctly distrust and override, and an overridden recommendation adds a workflow step rather than replacing one. Carriers with the longest legacy of separate systems by line and region, often the largest commercial writers, face the highest remediation cost before reliable models can be trained at scale.
Strategy fragmentation compounds it. Without a capability map, tools get bought in response to vendor demonstrations, which produces exactly the point-solution accumulation that delivers narrow gains and no systemic movement in loss or expense outcomes.
Training is the clearest reading of the 70%. The 12% who train regularly are the carriers whose underwriters and claims handlers can evaluate an AI output, calibrate when to accept it, and redesign their own work around it. Deploying without that is funding the 30% and leaving the 70% unaddressed.
The timing makes the gap harder to close rather than easier. BCG projects 22% of senior underwriters will have retired by 2026, and a carrier that has not redesigned its workflows loses that judgment into a market paying full rate to replace it. A carrier that has depends less on individual underwriter knowledge, because AI-assisted triage and rules-based clearance carry the standard risks and human review concentrates where the model flags complexity. The same attrition wave is a staffing problem for one group and a non-event for the other.
Further Reading on actuary.info
- BCG's AI-First P&C Insurer Blueprint Sequences Transformation in Three Phases - The Deploy-Reshape-Invent framework in detail, $35-60B cost reduction projections, and actuarial implications for pricing, reserving, and model validation workflows.
- Insurer AI Adoption Hits 82% But Only 7% Reach Full Scale - Sedgwick claims-specific data quantifying the vendor fragmentation, data silos, and NAIC regulatory overlay that block scale in claims operations specifically.
- Travelers Q1 2026: Separating the Tech Signal from the Cat-Loss Swing - How Travelers' dozens of scaled generative AI tools appear in expense ratios and LAE versus the headline combined ratio improvement driven by cat normalization.
- Progressive's Record Media Spend and the ML Pricing Edge Behind It - How proprietary pricing model precision drives the growth strategy the WTW analytics-leader combined ratio data reflects at the carrier level.
- Inside the Covenir Workforce Readiness Gap - Survey data on the training deficit behind the adoption-scale gap, with headcount and budget findings that map directly to the 12% of carriers regularly investing in analytics skills.