Understanding Ai Underwriting Tools 2026 for Insurance Brokers

AI underwriting tools 2026 now process standard commercial insurance submissions in under 3 minutes, with data extraction accuracy between 92% and 97%. The market has moved decisively past the pilot stage. According to McKinsey 2025, 68% of the top 50 P&C carriers now use AI-assisted underwriting in production environments, up from 30% in 2023. For brokers, this shift means faster quotes, more consistent pricing on standard risks, and a fundamental change in which skills drive revenue at the agency level.

This guide compares the six leading AI underwriting platforms, breaks down what each tool does, who has adopted it, and what brokers can expect when working with accounts processed by each system.

Key Takeaways

• McKinsey 2025 reports that 68% of the top 50 P&C carriers now use AI underwriting in production, up from 30% in 2023
• Cytora reduces commercial submission triage time by up to 70%, according to carrier case studies published in 2025
• Planck Data generates 200+ data points per account from public sources before a human underwriter reviews the file
• Cape Analytics processes over 100 million property locations annually using aerial imagery and machine learning (Cape Analytics 2025)
• Friss detects fraud signals in under 60 seconds, reducing claims leakage by 17% on average for adopting carriers (Friss 2025)
• BOP, personal auto, and small-account workers comp represent the three lines with highest AI underwriting adoption rates in 2026 (AM Best 2025)

Why AI Underwriting Tools Expanded So Fast in 2025-2026

The acceleration had two causes. First, carrier combined ratios worsened through 2023 and 2024, pushing leadership teams to reduce underwriting expenses per policy. Second, large language models matured enough to read unstructured submission documents, extract ACORD fields, and flag missing data without human intervention.

Accenture 2025 estimates that AI underwriting tools reduce the average underwriting cost per commercial policy by 22% when fully deployed. Carriers passed part of that savings into faster turnaround times. Brokers on the receiving end now compete on speed as much as coverage quality.

The shift also changed what a submission needs to contain. AI tools penalize incomplete submissions more harshly than human underwriters did. A missing loss run or an incomplete schedule of values no longer just delays the quote. It pushes the account to a referral queue that can add 3 to 5 business days.

Platform 1: Cytora

Cytora is a commercial lines underwriting workflow platform used by carriers including Zurich, RSA, and Hiscox. The platform ingests email submissions, extracts structured data from attachments, scores each risk against the carrier's appetite, and routes files to the correct underwriter or automated decision engine.

Cytora does not build its own pricing models. Instead, it sits on top of a carrier's existing rating engine and improves the intake and triage layer. Carrier data published in 2025 shows that Cytora reduces average submission triage time by 60% to 70%.

For brokers, Cytora-enabled carriers typically return acknowledgment faster and request fewer follow-up data points because the AI has already pre-populated what it could find in the submission package. The practical implication: complete submissions submitted to Cytora-enabled carriers move to quote faster than to carriers using manual triage.

Pricing for Cytora is carrier-side. Brokers do not pay for it directly. Integration with carrier portals and APIs is available for agencies with sufficient submission volume.

Platform 2: Federato

Federato focuses on underwriting decision support for complex commercial lines. The platform uses reinforcement learning to show underwriters which risks in their queue are highest value given current portfolio exposure, pricing conditions, and appetite.

Carriers using Federato include Markel and several specialty MGAs. Unlike Cytora, Federato is designed for accounts that still require human judgment, but it guides the underwriter toward faster, more consistent decisions. According to Federato 2025 documentation, carriers using the platform improve underwriting profitability by 4 to 8 percentage points on combined ratio over two years.

For brokers working on E&S and specialty accounts, Federato-enabled carriers tend to give more detailed declination reasons and more predictable timelines. The platform tracks which submission characteristics correlate with declines and surfaces that information to the underwriter at quote time.

Brokers cannot submit directly through Federato. It operates inside the carrier's system. The broker experience is improved response quality and consistency rather than a direct workflow change.

Platform 3: Planck Data

Planck Data is an AI data enrichment platform. When a broker submits an account, Planck automatically pulls 200+ data points about the risk from public web sources, business databases, review platforms, and regulatory filings. This data populates underwriting worksheets without the agent or underwriter manually gathering it.

Carriers using Planck include Markel, Westfield, and Employers. Planck 2025 reports that its platform reduces underwriter data-gathering time by 45% per account. For standard BOP accounts, Planck can assemble enough information to return a quote without a supplemental application in many cases.

The broker implication is significant. Planck-enabled carriers expect less supplemental data from agencies because the AI has already gathered it. However, if the broker's client has inaccurate information on the web (wrong SIC code, outdated revenue figures, incorrect employee count), Planck may flag discrepancies that require explanation. Keeping client data accurate in public sources becomes a subtle but real part of submission management.

Planck charges carriers on a per-enriched-account basis. Brokers do not pay directly but benefit from faster decisions.

Platform 4: Cape Analytics

Cape Analytics specializes in property underwriting. The platform uses aerial imagery, satellite data, and computer vision to assess roof condition, construction type, secondary structures, and property hazards without a physical inspection.

Cape Analytics 2025 reports that the platform processes over 100 million property locations annually. Carriers including Hippo, Nationwide, and Openly use Cape Analytics to pre-bind property risk assessment. For standard homeowners and small commercial property accounts, Cape Analytics can eliminate the need for an inspection entirely.

For brokers, this means that property submissions to Cape Analytics-enabled carriers process significantly faster. The tradeoff: the AI's property assessment can differ from what the insured reports. If the roof condition detected by aerial imagery does not match the application, the carrier may adjust the quote or request documentation.

Brokers handling commercial property should verify that the insured's stated property conditions match what is visually apparent from satellite view before submitting. Tools like Google Maps satellite view provide a rough preview of what the carrier AI will see.

Platform 5: Friss

Friss is an AI fraud detection and risk scoring platform used primarily during the quote-and-bind process. The platform analyzes new applications for fraud signals, claims history indicators, behavioral anomalies, and network connections to known fraud cases. Friss 2025 reports average claims leakage reduction of 17% for carriers that deploy the platform across personal and commercial lines.

Carriers using Friss include several European and US mid-market insurers. For legitimate accounts, Friss is invisible: the submission flows through normally. For flagged accounts, the carrier may request additional documentation or decline to quote.

From the broker's perspective, Friss occasionally flags accounts that appear suspicious based on data patterns even when the risk is legitimate. A business located at an address associated with prior fraud claims, or a client whose name matches a flagged entity, may receive a manual review request. When that happens, providing proactive documentation (financial statements, prior carrier references, loss run commentary) moves the account through faster.

Friss charges carriers on a subscription plus per-transaction basis. Broker-direct access is not available.

Platform 6: Bdeo

Bdeo is a visual intelligence platform used in property and auto underwriting and claims. The platform analyzes photos and videos submitted by insureds or agents to assess property condition, vehicle damage, and construction quality. Bdeo's technology enables carriers to conduct remote property inspections without sending an adjuster or inspector to the site.

Bdeo 2025 reports that remote visual inspections using the platform take an average of 4 minutes per property compared to 3 to 5 business days for a traditional inspection appointment. Carriers using Bdeo can accelerate both the underwriting and claims processes for property and auto lines.

For brokers, Bdeo-enabled carriers may ask the agent or insured to submit photos through a carrier app or portal link during the application process. Agents who help clients submit high-quality photo documentation up front see fewer inspection delays and faster policy issuance.

AI Underwriting Tool Comparison by Line of Business

AM Best 2025 notes that personal auto has the highest AI adoption because the data sources (motor vehicle records, telematics, credit proxies) are the most structured and most available. BOP ranks second because small business risks fit within tight appetite parameters that AI can evaluate consistently.

E&S and specialty lines remain largely human-underwritten because the risk characteristics fall outside the training data ranges that AI systems are calibrated on. Expect that share to grow as specialty carrier AI training data accumulates over 2026 and 2027.

How Agents Interact with AI-Underwritten Accounts

Working with AI-underwritten accounts requires a different preparation process than working with traditional underwriters. Here is what changes.

First, submission completeness becomes non-negotiable. AI systems route incomplete submissions to referral queues automatically. A missing loss run or a blank field on an ACORD application that a human underwriter might overlook and request later will stall an AI-processed submission for days.

Second, supplemental applications matter less for AI-enabled carriers on standard lines. Planck and similar enrichment tools pre-fill most of the data that supplemental applications used to gather. Sending a supplemental where one is not required slows the process.

Third, data accuracy matters more than it used to. AI tools cross-reference submitted information against public data sources. Discrepancies between what the client reports and what public records show generate flags that require manual resolution.

Gartner 2025 research indicates that agents who adapt their submission preparation processes to AI-carrier requirements reduce their average quote turnaround by 2.3 business days per account compared to agents submitting without adjustment.

Which Lines Have the Highest AI Underwriting Adoption in 2026

According to AM Best 2025, the three lines with the highest AI underwriting adoption are personal auto (81% of carriers using AI in some form), BOP for small commercial accounts (74%), and personal property/homeowners (68%). Workers comp for accounts under $25,000 in premium follows at 61%.

The common thread is data availability. Lines where structured, third-party verifiable data exists (driving records, property registries, payroll reports) are the easiest to automate. Lines where judgment plays a larger role (construction, hospitality, professional liability for complex firms) remain lower-adoption.

For agency strategy, this means that book composition matters when evaluating which AI tool investments to prioritize. An agency with 60% BOP and personal lines will benefit more immediately from learning Cytora- and Planck-enabled carrier workflows than an agency focused on E&S casualty.

Pricing and Integration Requirements by Platform

Cytora: Carrier-side deployment. No direct cost to brokers. API integration available for high-volume agencies through carrier portal programs.

Federato: Carrier-side deployment. Broker access is indirect. Integration points are through carrier portals and email submission workflows already in use.

Planck Data: Carrier-side, per-account enrichment pricing. Brokers benefit at no cost. Agencies can access Planck data directly through select MGA and wholesaler partnerships at per-enrichment rates.

Cape Analytics: Carrier-side for property lines. Some carriers expose the Cape score to brokers through their portal. No direct broker-purchase model as of April 2026.

Friss: Carrier-side subscription. Not available for direct broker purchase. The broker interaction is reactive: responding to flags rather than accessing the platform.

Bdeo: Carrier-side and MGA-side. Some carriers provide a Bdeo-powered photo submission link to agents as part of the new business workflow. Cost is carrier-borne.

What Brokers Should Expect Over the Next 12 Months

McKinsey 2025 projects that the share of carriers using AI underwriting in production will reach 80% among the top 50 by the end of 2026. The growth will be concentrated in two areas: broader rollout of existing pilots to full books of business, and expansion into lines that are currently at lower adoption rates.

For brokers, the practical shift over the next 12 months will be a higher proportion of submissions that receive instant or near-instant decisions. Accounts that once waited 3 days for a quote will receive quotes in under an hour. That speed changes how clients make decisions and how agents manage pipelines.

Applied Systems 2025 data shows that agencies connecting their AMS directly to AI-enabled carrier portals process 40% more submissions per producer per month than agencies still using manual submission methods. The gap between connected and disconnected agencies will widen as AI adoption expands.

Frequently Asked Questions

What are the leading AI underwriting tools used by P&C carriers in 2026?

The six leading platforms are Cytora (commercial lines triage), Federato (complex lines decision support), Planck Data (data enrichment), Cape Analytics (property imaging), Friss (fraud detection), and Bdeo (visual inspection). Most carriers combine more than one of these tools rather than using a single platform.

How does AI underwriting affect broker submission requirements?

AI-processed submissions require higher completeness upfront. Missing fields, absent loss runs, or inconsistent data between what the agent submits and what public records show will trigger manual referral queues. Brokers who prepare complete, accurate submissions see faster decisions on AI-enabled carriers.

Which lines of business have the highest AI underwriting adoption in 2026?

Personal auto leads at 81% carrier adoption, followed by BOP at 74% and personal property at 68%, according to AM Best 2025. Workers comp for small accounts is at 61%. E&S and specialty lines remain the lowest at 28%.

Do brokers pay for any of these AI underwriting platforms?

All six platforms described in this guide operate as carrier-side technology. Brokers do not pay for Cytora, Federato, Planck, Cape Analytics, Friss, or Bdeo directly. The cost is absorbed by the carrier. Some platforms offer direct broker or MGA access at per-transaction pricing for agencies that want to enrich their own submission data before sending to carriers.

How accurate are AI underwriting decisions compared to human underwriters?

Accuracy varies by line and platform. For BOP and personal auto on standard risks, AI systems reach 92% to 97% data extraction accuracy (McKinsey 2025). For complex commercial lines, AI tools assist rather than replace human judgment, which makes direct accuracy comparison less meaningful. Carriers report that AI-flagged risks have lower loss ratios on average than risks that bypassed the AI triage layer.

What should agents do when a submission is flagged by a carrier's AI system?

When a carrier AI flags an account for manual review, the fastest resolution path is proactive documentation. Provide financial statements, a loss run narrative, prior carrier references, and any context that explains data discrepancies. Do not wait for the underwriter to request documents one at a time. Sending everything upfront typically cuts manual review time by 50% or more.

Improve your submission workflow →

Written by Javier Sanz, Founder of BrokerageAudit. Last updated April 2026.