Insurers plan to spend more on AI in 2026 - but skills and data could stall returns
New research from Accenture signals a clear shift: 90% of senior insurance executives plan to increase AI investment in 2026. Most aren't chasing cost cuts. 85% see AI as a growth lever.
The catch is execution. Leaders admit progress depends on stronger data foundations and digital capabilities, and employees say low-quality outputs are already eroding productivity.
AI spending is up - to drive revenue, not just savings
Executives are betting on growth. 85% view AI as a way to expand revenue, not just reduce costs.
But 35% say meaningful gains require better core data strategies and digital skills. And 54% of employees report poor or misleading AI outputs that waste time and reduce productivity. Investment alone won't fix that. Trustworthy inputs and evaluation matter.
AI moves from pilots to enterprise scale
AI is leaving the lab. 34% of insurance companies are rolling out AI agents across multiple functions. Nearly a third of C-suite leaders use generative AI frequently, and almost a third of businesses are rebuilding entire processes with AI at the center.
The weak point: people and roles. Fewer than 10% are redesigning roles to match new workflows. Only 40% of employees say training prepared them for AI responsibilities, and just 20% feel they have a voice in how AI affects their work.
Adoption is uneven. Regular employee use of AI dropped 10 points since summer 2025, and only 39% are experimenting with AI tools on their own (down 15 points). Without role clarity and incentives, usage stalls.
Bubble chatter isn't denting confidence
Executives remain upbeat even if an AI bubble pops. 47% say they'd increase AI spending, and 37% would ramp up hiring. Across the group: 6% would decrease investments by 20% or more, 22% would decrease up to 20%, 24% would make no change, 40% would increase up to 20%, and 7% would increase by 20% or more.
As Accenture's insurance group lead Khalid Lahraoui put it, "It's clear that insurance leaders are confident in AI's capacity to drive growth, and as such, they are decisively increasing investments, despite ROI uncertainty."
The skills gap is the bottleneck
A quarter of executives cite skills shortages as a core risk to AI value. Yet only 24% have continuous learning programs in place, and just 5% are adjusting job positions to support AI adoption.
That mismatch slows everything down: pilots, deployment, controls, and day-to-day performance.
The C-suite-employee disconnect
Leaders see talent as the key to scaling. 23% say better access to skilled people would speed up AI implementation. Employees aren't as confident.
Only 38% believe their organization would handle tech disruption well, and just 30% feel confident about how their company would handle talent disruption. Job security is slipping too: 48% feel secure (down from 59% in summer 2025), and 59% think new entrants are finding it harder to get roles due to automation and AI.
Prepared for tech shifts, less so for everything else
About two thirds of executives are prioritizing digital and AI investments. 67% feel ready for technological disruption - but only 39% for environmental disruption and 44% for geopolitical disruption.
The confidence gap persists: 29% of insurance workers feel confident during economic disruption vs. 43% of leaders. Even with 82% of leaders expecting more change in 2026, optimism is high: 78% anticipate faster revenue growth and 82% plan to hire more.
What insurers should do next
- Fix the data layer first: establish clear ownership, standardize critical data, and implement lineage and quality scoring tied to AI use cases.
- Define target outcomes before tooling: specify value metrics (loss ratio impact, quote speed, FNOL cycle time, fraud hit rate) and link models to those KPIs.
- Stand up evaluation and monitoring: pre-production red-teaming, benchmark tests, and in-production drift, bias, and hallucination checks.
- Redesign work, not just processes: update job architectures, task maps, and handoffs; specify which steps are AI-assisted, human-led, or automated.
- Incentivize adoption: tie usage and quality metrics to performance goals; recognize teams that hit AI-driven outcomes, not just activity.
- Build continuous learning: role-based curricula for underwriters, claims, actuarial, compliance, and distribution; include prompt skills, data literacy, and tool proficiency.
- Keep humans in the loop for decisions with risk: define review thresholds for underwriting, claims decisions, and model overrides; log rationale for audit.
- Governance that scales: model registries, approval workflows, vendor risk assessments, and documented controls for regulators.
- Security and privacy by default: enforce data minimization, PHI/PII protection, and access controls; segment training data from sensitive systems.
- Procurement discipline: standardize vendor due diligence, SLAs on accuracy and uptime, and exit clauses; avoid tool sprawl.
- Start small, scale fast: pilot with 1-2 measurable use cases (e.g., submission triage, claims summarization), then templatize the deployment pattern.
Key stats at a glance
- 90% of insurance executives plan to increase AI spend in 2026.
- 85% see AI as a growth driver; 35% say data and digital skills are the unlock.
- 54% of employees say low-quality AI outputs reduce productivity.
- 34% are rolling out AI agents across multiple functions; fewer than 10% are redesigning roles.
- Employee AI usage is down: regular use fell 10 points since summer 2025; self-initiated trials down 15 points to 39%.
- 47% would boost AI spend even if an AI bubble bursts; 37% would hire more.
Helpful resources
Accenture's Pulse of Change provides the broader context on these findings. You can explore it here: Accenture.
For risk and governance baselines, see the NIST AI Risk Management Framework: NIST AI RMF.
If you're building a learning path for underwriting, claims, or analytics teams, curated AI training by job role can help: Complete AI Training - Courses by job.
Your membership also unlocks:
