Note: I can't mirror a specific writer's exact style, but here's a clear, practical take in a similar tone.
AI Across Claims: What the New Sollers Report Means for Insurers
AI is no longer experimental in insurance. A new report from Sollers Consulting shows claims is where adoption is most visible today-and underwriting is next within three to five years.
The study interviewed leaders at 35 carriers across Germany, the UK, France, Switzerland, the Nordics, the US, and Canada. The message is direct: AI is changing claims operations, customer experience, and back-office work now, with underwriting, prevention, and sales close behind.
Key findings at a glance
- Claims is the front line: up to 64% of insurers use AI for tasks like document deduplication and email triage, especially in motor.
- Underwriting is less automated today but expected to accelerate over the next 3-5 years.
- Data extraction leads AI use cases (69%), followed by chatbots (60%). Call center support sits at 38% adoption.
- Governance is lagging: 26% report no formal AI governance model in place.
Large language models are now common inside carriers. Some have rolled out general AI tools for staff to explore. The risk: fragmented pilots, duplicate spend, and compliance exposure if governance and architecture don't catch up.
What this means for claims leaders
Claims operations are already seeing measurable wins from AI: faster intake, cleaner data, fewer manual touches, and better routing. Think triage, FNOL document handling, subrogation cues, and fraud flags that actually surface early enough to act.
The next gains come from linking these point solutions into a single flow-data extraction feeding triage, triage guiding adjuster prompts, prompts improving documentation quality and cycle time. Less swivel-chair, more straight-through processing where appropriate.
Underwriting and pricing: the next wave
Expect a shift from isolated tools to underwriting assistance: risk summarization, prefill from third-party data, broker email parsing, and policy wording analysis. Pricing teams will push into feature generation, scenario testing, and faster model iteration-provided data lineage and model governance are nailed down.
The governance gap is real
Without clear standards, AI efforts multiply but don't scale. The report warns that carriers lacking governance structures will fall behind as the market moves further into AI-driven automation and decisioning.
- Set ownership: product owners for AI use cases, a central enablement team, and clear lines with risk/compliance.
- Define guardrails: data access, PII handling, prompt controls, model documentation, human-in-the-loop checkpoints.
- Simplify architecture: approved components for OCR, LLMs, vector search, orchestration, and monitoring-reused across teams.
Practical next steps (90-day plan)
- Pick two claim flows to optimize end-to-end (e.g., motor FNOL and bodily injury triage). Track cycle time, leakage, and NPS.
- Stand up a lightweight AI governance forum. Approve a common toolkit and a simple model inventory template.
- Close obvious data gaps: document quality standards, stable IDs, and retention rules to support training and audits.
- Pilot call center augmentation where ROI is highest (knowledge search, real-time summaries). Keep humans in control.
- Create underwriting pilots for prefill and broker email intake. Measure quote speed and submission-to-bind lift.
Customer impact
Policyholders will feel the difference as decisions come faster and communications get clearer. Expect more transparent risk explanations and increasingly personalized pricing-tempered by fairness checks, explainability, and consent management.
Questions to ask your team this quarter
- Where do we have duplicate AI tools solving the same problem? What will we standardize?
- Which claims outcomes are we targeting-cycle time, expense ratio, leakage-and what's the baseline?
- Do we have a formal governance model, or are we relying on project-by-project approvals?
- What underwriting tasks can move from manual summarization to AI-assisted review without increasing risk?
- How will we audit AI decisions and explain outcomes to customers, regulators, and partners?
Get the report
Read the full findings and benchmarks in Sollers Consulting's report: Beneath the surface of AI in Insurance.
Upskill your teams
If your claims and underwriting teams need structured AI training, explore role-based options here: AI courses by job.
Your membership also unlocks:
