Sixfold
by Sixfold
An AI underwriter that learns your book and takes submissions through to bind-ready
Sixfold is an AI underwriting platform for insurance carriers that reads inbound submissions, scores risk against the carrier's own appetite, and can run straight through to quote- and bind-ready output. Each carrier gets a permanently isolated model instance that keeps learning from its own underwriters' decisions, so recommendations reflect that specific book rather than generic guidelines.
Sixfold builds AI for insurance underwriting, and in June 2026 launched AI Underwriter, an agent that takes a submission from intake through to quote- and bind-ready materials. The architecture has two layers. Underneath sits what the company calls an Underwriting Brain: a foundation model trained on professional underwriting credentials and reasoning patterns spanning more than 40 lines of business. On top of it, each carrier gets a permanently isolated instance that continuously learns from that carrier's own submissions, underwriter overrides and outcomes, which is how Sixfold captures appetite that is never written down in any guideline document. In practice the agent reads complex documents such as loss runs and statements of value, extracts the data and flags what is missing, scores risk with explainable signals that cite their sources, tests portfolio fit, and recommends the next action on every submission. Carriers choose one of three autonomy levels: recommendation only, quote-ready for underwriter approval, or straight-through processing that produces bind-ready output with no human review for defined categories. It integrates with existing underwriting workbenches and policy administration systems rather than replacing them, and covers both property and casualty and life and health, where it surfaces impairments from prescription histories and lab results. Sixfold has processed over a million submissions for insurers representing roughly $265 billion in gross written premium, and AI Underwriter went live across six carriers representing about $270 billion. Named customers include Zurich North America, Guardian, AXIS, Generali GC&C and Skyward Specialty. Founded in 2023 in New York by Alex Schmelkin, Jane Tran and Brian Moseley, it has raised roughly $52 million, most recently a $30 million Series B led by Brewer Lane in January 2026 with strategic investment from Guidewire.
The chief underwriting officer or head of underwriting operations at a P&C or life carrier whose underwriters spend hours per submission on document review before any risk judgement happens.
Submissions arrive pre-read, pre-scored against your own appetite and ranked by portfolio fit, with defined categories able to run straight through to bind-ready.
At a Glance
- Category
- Industry & Government
- Pricing
- Contact for pricing
- Target Market
- Chief Underwriting Officers, CIOs, Heads of Underwriting Operations, Insurance COOs, Actuaries
- Deployment
- Cloud-only
- Founded
- 2023
- Headquarters
- New York, United States
- Customers
- Six P&C carriers live on AI Underwriter representing roughly $270B in gross written premium (June 2026); 1M+ submissions processed across 40+ lines of business
Key Features
- ✓Underwriting Brain foundation model
Pre-trained on professional underwriting credentials and reasoning across 40-plus lines, so it is not learning the domain from scratch.
- ✓Carrier-isolated learning instance
Each carrier's instance trains only on its own decisions and overrides, keeping data separated and capturing appetite that no guideline documents.
- ✓Complex document ingestion
Parses loss runs, statements of value and full submission packets, extracts the structured data and flags what is missing.
- ✓Explainable risk signals with citations
Every score links back to the source document, which is what makes the recommendation defensible under audit and model risk review.
- ✓Three configurable autonomy levels
Carriers pick recommendation-only, quote-ready for underwriter approval, or straight-through bind-ready output per business category.
- ✓Workbench and policy admin integration
Slots into existing underwriting workbenches and policy administration systems instead of forcing a core platform replacement.
- ✓Annual Responsible AI Report
A published report covering governance, fairness and transparency, aimed squarely at regulator and model risk scrutiny.
Capabilities
Use Cases
- •Submission triage and appetite scoring
Inbound submissions are scored against the carrier's real appetite so underwriters open the ones most likely to bind first.
- •Quote turnaround acceleration
Skyward Specialty cut average quote response time by 35% across eleven underwriting teams using the platform.
- •Middle-market underwriter leverage
Zurich North America reports saving up to two hours per submission across more than 200 middle-market underwriters.
- •Straight-through processing
Defined low-complexity categories move from submission to bind-ready materials with no underwriter review in the loop at all.
- •Life and health case analysis
Surfaces impairments from prescription histories and lab results, with recommendations aligned to the carrier's own underwriting guidelines.
Ideal For
Best For
- ✓P&C carriers with high submission volume where underwriters lose hours to intake and document review
- ✓Life and health carriers automating impairment review from prescription histories and lab results
- ✓Specialty and middle-market lines where appetite is nuanced and lives in underwriter judgement, not guidelines
- ✓Carriers that want to keep their existing underwriting workbench and policy administration stack in place
- ✓Insurers needing explainable, source-cited output to satisfy model risk and regulatory review
Not Ideal For
- ✗Anyone outside insurance — the model, the training data and the entire workflow are built around underwriting
- ✗Carriers that cannot sustain the 90%+ underwriter adoption the continuous learning loop depends on to stay current
- ✗Buyers who require published compliance certifications up front: Sixfold's security page documents AWS infrastructure, quarterly vulnerability scans, annual penetration testing and encryption, but names no SOC 2 or ISO certification
- ✗Carriers unwilling to accept the vendor lock-in that carrier-specific fine-tuning accumulates quarter over quarter
- ✗Small carriers or MGAs wanting a light per-deal copilot rather than a deep platform integration
Deployment
Market & Ratings
Six P&C carriers live on AI Underwriter representing roughly $270B in gross written premium (June 2026); 1M+ submissions processed across 40+ lines of business
Market Analysis
Pros
- ✓Domain pre-training rather than a general LLM with an insurance prompt, which shortens the path to useful output
- ✓Per-carrier isolation delivers both data separation and appetite capture that a shared generic model cannot replicate
- ✓Deployed at named tier-one carriers (Zurich North America, Guardian, AXIS, Generali GC&C, Skyward Specialty) with specific published time savings
- ✓Guidewire's strategic investment signals alignment with a core policy administration incumbent
- ✓The explicit three-level autonomy ladder lets a carrier start advisory and escalate rather than committing to autonomous bind on day one
Cons
- ✗The 1.5 million submissions behind the models span the hard market of 2024-2025, so an independent actuarial analysis warns the system may recommend tighter terms than a softening market requires
- ✗The claimed 15%+ hit-ratio gain may reflect the model steering flow toward preferred business rather than genuine risk selection — the loss ratio would only reveal that several quarters later
- ✗Appetite drift: when carrier strategy shifts, retraining requires labelled examples of the new appetite, because the model learned from decisions rather than from the guideline documents
- ✗Carrier-specific fine-tuning creates switching costs that accumulate every quarter, a real lock-in risk flagged in independent analysis
- ✗Under ASOP No. 56 the carrier retains the validation duty when the system binds autonomously, demanding transaction-level auditability and pricing consistent with filed rates
- ✗Pre-existing coding quirks in a carrier's historical data become training signal, potentially encoding workflow artefacts as appetite preferences that never existed
- ✗No SOC 2, ISO 27001 or ISO 42001 certification is named on the public security page, only infrastructure practices
Pricing
Enterprise
Contact for pricing
- ✓Carrier-isolated fine-tuned instance
- ✓AI Underwriter with configurable autonomy level
- ✓Workbench and policy administration integration
- ✓P&C and/or life and health lines
Sixfold publishes no pricing at all; deals are enterprise sales-led and scoped per carrier, because every customer receives a fine-tuned isolated instance integrated with its own workbench and policy administration systems. Independent analysis contrasts this with lighter per-deal copilots such as nsur.ai at roughly $2 per deal, but Sixfold's own figures remain unpublished.
Security & Compliance
Sources
This page was written from 6 sources, 4 on domains other than sixfold.ai.
- 1.sixfold.ai — sixfold.aivendor
- 2.sixfold.ai — securityvendor
- 3.fintech.global — sixfold launches ai underwriter for pc insurers
- 4.fintech.global — insurtech firm sixfold secures 30m to advance ai underwritin
- 5.actuary.info — sixfold ai underwriter institutional memory stp 2026
- 6.natlawreview.com — sixfold raises 30 million series b build ai underwriter
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Casepoint IQ
Governed agentic AI for eDiscovery, FOIA and investigations, on FedRAMP High and DoD IL6 infrastructure
Aderant Agent Center
Purpose-built AI agents for law firm back-office operations, on the Stridyn platform
DISCO Advanced Research
Agentic deep research for litigation — autonomous multi-step fact investigation with an auditable reasoning trail
Dili
AI-powered prevailing wage and certified payroll compliance for federally funded infrastructure projects