Insurance CEOs Put AI at the Center of 2026 Strategy
A new KPMG Insurance CEO Outlook signals a clear shift: AI is no longer a side project. About 73% of insurance leaders rank AI as a top investment priority, with roughly 67% planning to commit 10-20% of budgets to AI initiatives. Most expect measurable returns in one to three years.
The message is simple-AI is moving from tests to daily operations. The winners will be the carriers that execute cleanly, measure aggressively and upskill fast.
Where AI Is Delivering Now
- Claims: Faster validation, quicker approvals and automated payouts cut cycle times and errors. Photo analysis can auto-approve straightforward motor claims.
- Underwriting: Algorithms tighten risk assessment and pricing using live data, lifting price adequacy while keeping premiums fair.
- Customer service and ops: Routine work gets automated, response times improve, and teams focus on higher-value tasks.
What This Means for Your P&L
- Lower loss adjustment expense through straight-through processing.
- Improved combined ratio via sharper risk selection and pricing.
- Higher retention and NPS with faster answers and cleaner experiences.
Kenya: Fast Gains on Real Pain Points
Stronger AI adoption can reduce slow claims handling and clamp down on fraud-two long-standing friction points in the market. Local carriers are already using automation to cut turnaround times and tighten checks.
Progress will track industry commitment and supportive policy. Digital infrastructure and workforce skills need investment for benefits to hold and scale.
Risk, Regulation and Trust
About 77% of CEOs say slow regulatory progress could hold back results. Clarity on AI ethics, data protection and compliance remains patchy, and that can erode public trust if mishandled.
Don't wait. Adopt internal standards now-model risk controls, human-in-the-loop on high-impact decisions, data lineage, audit trails and bias testing. Frameworks like the NIST AI Risk Management Framework are a solid starting point. For context on sentiment, see KPMG's CEO Outlook.
90-Day Plan: Prove Value Fast
- Pick 2-3 high-yield use cases: FNOL triage, fraud flags at intake, pricing enrichment for one product line.
- Define hard metrics: FNOL-to-payment time, straight-through rate, detection lift, quote-to-bind time.
- Stand up data pipelines with clear ownership and quality checks. No clean data, no lift.
- Put in simple guardrails: approval thresholds, escalation rules, monitoring dashboards and rollback paths.
12-Month Build: Scale Without Chaos
- Modernize your data stack-governed sources, feature stores, versioned models and CI/CD for ML.
- Create cross-functional squads (claims, underwriting, compliance, data science, IT) with a fixed backlog and stage-gates.
- Vendor governance: outcome-based contracts, bias and performance clauses, transparent logs, secure integrations.
- Model Ops: drift alerts, challenger models, periodic revalidation and clear decommission criteria.
Budgeting That Pays Back
- Allocate the 10-20% across three buckets: Run (automation), Grow (pricing, distribution), Transform (new products, embedded). Weight to quick wins first.
- Tie every workstream to a financial KPI and a customer KPI. Kill or scale based on evidence, not slide decks.
Skills: Upskill, Then Hire
- Train underwriters, claims and service teams on AI literacy, prompt practices and data quality basics. A practical path: role-based programs like AI courses by job.
- Augment with ML engineers, product managers and data stewards who can ship safely and repeatedly.
What Good Looks Like (Targets to Track)
- Claims: 30-60% faster cycle time; +5-15 pts straight-through processing; measurable fraud detection lift.
- Underwriting: Faster time-to-quote; improved hit ratio at target profit; better price adequacy and loss ratio by segment.
- Service/Back office: Lower handle time, higher first-contact resolution, reduced rework.
Bottom Line
AI spend is shifting from experiments to operations. The upside is real-faster returns, cleaner processes, better service-but it hinges on clear guardrails, strong data and people who know how to use the tools.
Pick focused use cases, measure hard outcomes and invest in skills. That's how you turn a budget line into a better combined ratio.
Your membership also unlocks:
