Federato Claims [What It Changes]

TL;DR

• Federato Claims is an end-to-end claims system that records coverage decisions, reserve changes, and adjuster reasoning as structured data and sends those signals back into underwriting, pricing, and product while the claim is still open, as of the August 4, 2026 launch covered by FinTech Global.

• Federato says claims can make up as much as 80% of an insurer's costs, yet older claims systems log the decision and stop. Treat the 80% line as the vendor's, reported in that same write-up, not as a regulator's combined-ratio table.

• The product is available now, either added onto Federato's existing underwriting, policy-administration, and billing platform or sold on its own to carriers and managing general agents (MGAs) that already handle claims in-house.

• A two-truck HVAC shop, a 10-person agency, or a solo clinic will not buy this core. They should still care about the pattern: the job already happened, the system of record is empty or closed, and the next price, form, or appetite decision never sees the file.

Key Takeaways

• Federato Claims is a claims-to-underwriting loop, not a bolt-on summarizer sitting on a legacy claims admin.

• Coverage calls, reserve movement, and adjuster rationale are the objects. FNOL, payments, and a broker/policyholder portal are in the same product, per Beinsure.

• State insurance departments still expect AI used in claims to obey unfair-claims and unfair-trade rules. The NAIC model bulletin on insurer AI is the reminder, not Federato's marketing page.

• The first honest test is one open claim: the coverage decision, the reserve change, and the reason sit in a record an underwriter can read before the file is paid and closed.

The answer in plain English

Federato Claims is a claims-management product that treats an open claim as the start of the next underwriting, pricing, appetite, or policy-form decision instead of the end of a file. As of August 4, 2026, that is the product Federato, an AI-native insurance-platform vendor, said it shipped for carriers and MGAs.

A two-truck HVAC company already lives a cheaper version of this failure: the tech finishes the capacitor swap, types a note in the van, and the owner never sees that the same part failed on three rooftops until the parts bill lands. A 10-person agency archives the last dispute and prices the next retainer anyway. A solo clinic finishes the visit and the fee schedule does not move. Carriers do it at book scale: the adjuster decides coverage, sets a reserve, and the underwriter finds out months later. Federato's argument is that the closed file is the constraint that broke. The product does not prove the next rate filing is correct. It proves there is a pipe.

That is why a small operator should read this page even if they will never log into Federato. The useful question is not "does AI read the claim." It is "who owns the coverage call when software says the book should tighten, and what evidence sits next to that call."

Who should use this page

This page is for a claims leader, chief underwriting officer, product actuary, or MGA operator who already runs commercial or specialty claims in-house, and for any agency principal being sold "AI that closes the underwriting loop."

Red flags: treating a structured rationale as a signed coverage opinion; turning on Federato Claims and a bolt-on summarizer as if they were one SKU; skipping a human review because the reserve movement looks consistent.

What shipped, and when

According to FinTech Global, Federato introduced Federato Claims on August 4, 2026 as an end-to-end claims management system built on the same technology foundation as its underwriting, policy administration, and billing tools. Federato Claims launched 4 August 2026. Those dates are the trade press's, matching the company's own Claims product page, which also timestamps an "Introducing Federato Claims" item on August 4, 2026.

According to FinTech Global, Federato says claims can make up as much as 80% of an insurer's costs, and that claims is often where new loss trends first become visible. Claims can consume 80% of insurer costs. Treat that share as Federato's, reported by the outlet, not as a figure from the Insurance Information Institute's industry overview.

Beinsure's August 5, 2026 write-up repeats the same launch and adds operational detail: an AI agent at first notice of loss (FNOL), policy-language checks as facts arrive, an audit trail of rationale, and a portal so brokers and policyholders can see claim stage without calling the adjuster. That is secondary coverage of the same product, not an independent loss-ratio study.

Federato co-founder and chief technology officer William Steenbergen, quoted in the FinTech Global piece, said claims is the one place in the policy lifecycle where you find out if your underwriting was right, and that Federato Claims is built to treat that discovery as the start of the next decision: the reserve you adjust, the appetite you tighten, the policy form you fix before the next renewal.

Sources: FinTech Global; Beinsure; Federato Claims.

Why claims is the constraint that broke

The industry already spends most of the premium dollar on losses and loss-adjustment expense. According to the Insurance Information Institute, U.S. property/casualty net premiums written reached $857.8 billion in 2023, up 10.2% from 2022. P/C net premiums written hit $857.8 billion. That table is NAIC data sourced from S&P Global Market Intelligence, published by III.

According to the same III overview, P/C insurers incurred $627.4 billion of losses and loss-adjustment expenses in 2023 against $821.5 billion of premiums earned, and booked a $20.2 billion net underwriting loss. Those are calendar-year industry totals. They are not Federato's book.

The people who work the files are a large occupation. According to the Bureau of Labor Statistics Occupational Outlook Handbook, claims adjusters, appraisers, examiners, and investigators held about 389,700 jobs in 2025, with 2025 median pay of $78,020, and employment projected to decline 6% from 2025 to 2035. Adjuster jobs numbered 389,700 in 2025. The BLS NAICS 524 glance lists 241,680 claims adjusters inside that subsector and 97,600 underwriters; those are different scopes.

NAIC's insurance-fraud topic page, last updated June 23, 2026, is the other pressure on the same files. Cornell's Legal Information Institute note on insurance fraud distinguishes hard fraud from soft fraud, last reviewed June 2023. Federato Claims is not a fraud bureau. It says it will surface inconsistent decisions, repeated clause disputes, and clusters of reserve overrides.

Flood is a reminder that some claims never live only inside a carrier's admin. According to FEMA's flood-insurance page, NFIP is delivered by more than 47 companies plus NFIP Direct, covers 4.7 million policyholders, and provides nearly $1.3 trillion in flood coverage. NFIP covers 4.7 million policyholders. Federato does not rewrite that federal constraint.

Sources: Insurance Information Institute industry overview (NAIC data via S&P Global Market Intelligence).

How Federato Claims works

The mechanism is a shared architecture. Federato says Claims runs on the same AI-native stack and federated context graph as underwriting, policy administration, and billing, so a claim stays linked to the risk and the policy. Federato's Claims page describes capturing rationale as work happens, then turning coverage decisions, disputes, overrides, and severity patterns into structured signals.

On a file, Beinsure and the product FAQ describe an AI agent at FNOL, policy-language checks as facts arrive, structured reserves and reasons, pattern flags for drift and repeated disputes, feedback into wording, appetite, and renewal rate, and a portal so brokers and policyholders do not call the desk for status. The FAQ says Claims covers the full lifecycle (FNOL through payment and closure) and is built for commercial and specialty carriers. This is not a personal-auto chatbot sold to a two-person agency.

Federato's homepage still sells the rest of the lifecycle: agentic quoting, a Control Tower for portfolio steering, billing, and named customer quotes. Those homepage percentages (a displayed 90% reduction in systems used, an 89% reduction in time to quote) are vendor marketing for underwriting workflow, not measured outcomes of Federato Claims. Do not paste them onto this SKU.

Who shipped it, and what the regulators already require

Federato is the vendor. William Steenbergen is the named co-founder and CTO on the launch coverage. The company homepage lists carriers, MGAs, MGA aggregators, and mutuals as buyer types, and quotes operators including a Nationwide CIO on agentic AI. Nationwide is a named customer voice on that page, not a claims-outcome study.

The regulatory overlay is older than the product. According to NAIC's artificial-intelligence topic page (last updated April 3, 2026), 88% of 193 auto insurers responding said they use, plan to use, or plan to explore AI/ML models. 88% of 193 auto insurers use or plan AI. According to the same NAIC AI page, that figure was 70% of 194 home insurers, 58% of 161 life companies, and 92% of 93 health insurers. P&C respondents already reported AI in claims for accident image analysis, estimating ultimate settlement values, and fraud detection. Federato Claims is arriving into a market that has already told regulators it is using models on claims files.

The NAIC Model Bulletin on the Use of Artificial Intelligence Systems by Insurers, adopted December 4, 2023, reminds licensed insurers that decisions or actions impacting consumers that are made or supported by AI systems must comply with all applicable insurance laws, including unfair trade practices and unfair claims settlement practices. It expects a written AIS Program covering governance, risk management, and internal audit, scaled to the degree of potential harm. It also names model drift. A claims system that feeds underwriting in real time is exactly the kind of AI system that bulletin is aimed at. As of March 2026, NAIC said an AI Systems Evaluation Tool was being piloted by 12 states.

NIST's AI Risk Management Framework is voluntary. NIST released AI RMF 1.0 on January 26, 2023 and NIST-AI-600-1 on July 26, 2024. A claims-to-pricing loop that a market-conduct examiner can ask about is the kind of system that playbook was written to document.

The U.S. Treasury Federal Insurance Office monitors the sector under Title V of Dodd-Frank. It does not license Federato. State departments do. CISA, HHS HIPAA, GDPR.eu, and the FTC are adjacent constraints. None of those pages mention Federato Claims.

Agencies that already compare Applied Epic vs Salesforce Financial Services Cloud or that are mapping insurance automation workflows will meet this product as another system that wants the same policy, claim, and producer context. Retention work in agency marketing workflows is downstream of the coverage call, not a substitute for it.

USTA analysis

USTA analysis (derived only from figures already cited above). Inputs: III 2023 P/C losses and LAE of $627.4 billion; III 2023 P/C premiums earned of $821.5 billion. Arithmetic: 627.4 ÷ 821.5 = 0.764, or a 76.4% calendar-year loss-and-LAE ratio. Federato's vendor line is that claims can make up as much as 80% of costs. Those are not the same object: III is losses plus LAE over earned premium for U.S. P/C in 2023; Federato is a company statement about "costs." The vendor ceiling sits next to an industry ratio already in the mid-70s. Federato did not publish a price. Do not invent one.

A second derived check, same sources: III 2023 other underwriting expenses of $213.9 billion plus losses and LAE of $627.4 billion plus policyholder dividends of $3.6 billion equals $844.9 billion of underwriting outgo against $821.5 billion earned, which is the $20.2 billion underwriting loss on that table. A claims system that only speeds correspondence does not change that identity. Federato has not published a combined-ratio delta. The 76.4% ratio is the industry baseline, not a Federato result.

Sources: III; FinTech Global; BLS OOH. Arithmetic is USTA's.

What it is not

It is not a state insurance department. It is not the NAIC bulletin. It is not a substitute for an unfair-claims-practices file. It is not FNOL-only. It is not a personal-lines agency management system. It is not a promise that loss ratio will fall. It is not the SBA disaster loan a small business files after a flood. It is not DOL wage-and-hour guidance for the adjuster workforce. Teams that already route claim documents through US Tech Automations workflows will still need a human coverage owner and a market-conduct file.

How a small shop should read the same loop

If you run a two-truck HVAC company, you will not implement Federato Claims. The analog is: every callback on the same unit should change the next maintenance price before the warranty expires. If you run a 10-person agency, every coverage dispute on a clause should change the endorsement you sell next week. If you run a solo clinic, every denied claim should change the eligibility check at intake, not the write-off at year-end.

US Tech Automations is the place a shop maps that analog when the insurance core is out of reach: document in, decision recorded, next price or form updated, human sign-off on the high-impact step.

Signal vs Speculation

Demonstrated (sourced): Federato announced Federato Claims on August 4, 2026 as an end-to-end claims system on the same architecture as underwriting, policy, and billing, available immediately as an add-on or standalone, per FinTech Global and Beinsure. William Steenbergen is the named executive. NAIC has surveyed AI use in claims and adopted a 2023 model bulletin. III's 2023 P/C table shows $627.4 billion of losses and LAE on $821.5 billion earned. BLS counts hundreds of thousands of adjuster jobs and a declining outlook.

Unknown: any independent combined-ratio, leakage, cycle-time, or customer-experience result attributable to Federato Claims. Any price. Any named carrier go-live on this SKU as of the pages we opened. Whether the federated context graph survives a market-conduct request for training data, third-party models, and adverse-consumer-outcome testing under the bulletin.

Our read: if the product does what the launch copy says, the 12-to-36-month effect on small and mid-size businesses is indirect. MGAs and specialty carriers that sell through independent agencies will ask those agencies for cleaner bordereaux and faster FNOL facts. HVAC and clinic operators will not buy Federato. They will buy the pattern: do not let the completed job die in a closed note. If carriers actually route open-claim signals into rate and form, independent shops should expect more questions at bind and fewer surprises at renewal. If they do not, this remains a vendor architecture story on top of a 76.4% industry loss-and-LAE ratio that did not need Federato to exist.

FAQs

What is Federato Claims?

Federato Claims is Federato's end-to-end claims management product that records coverage decisions, reserves, and adjuster reasoning as structured data and feeds those signals into underwriting, pricing, and product while the claim remains open, as described in the August 4, 2026 launch coverage.

Is Federato Claims available now?

Yes. FinTech Global and Federato's Claims page both say it is available immediately, either as an addition to Federato's existing platform or as a standalone system for carriers and MGAs that already manage claims internally.

Does it replace the adjuster?

No. The launch copy and the product FAQ describe AI that prepares intake, surfaces policy issues, and structures rationale for adjuster review. Coverage calls, payments, and high-impact actions still sit with the carrier's people and with state unfair-claims rules.

How is this different from bolt-on claims AI?

Federato's stated difference is architecture: Claims shares the same graph as underwriting, policy, and billing, so the file is not a separate system that only summarizes itself. Bolt-on tools, in the company's telling, speed tasks inside a legacy admin without capturing the reasoning or the book-level pattern. That is a vendor contrast, not a bake-off we ran.

What should a small agency actually do?

Keep FNOL structured, keep dispute clauses tagged, and keep the producer note attached to the policy so that if a carrier or MGA later asks for open-claim context you are not rebuilding the file from email. Map that path once if the documents already live in your shop.

Does NAIC require this product?

No. NAIC's 2023 AI bulletin requires responsible use of AI systems that affect consumers, including claims. It does not name Federato. Twelve states were piloting an evaluation tool as of March 2026, per NAIC's AI topic page.

What number should I not quote as Federato's result?

Do not quote III's $857.8 billion of P/C premium, BLS's 389,700 adjuster jobs, FEMA's 4.7 million NFIP policies, or NAIC's 88% auto-insurer AI survey as Federato outcomes. Those are industry and labor figures. Federato's own 80% "costs" line is a company statement.

Glossary

• Federato Claims: Federato's August 4, 2026 claims system that structures coverage, reserves, and rationale and feeds them back into underwriting while the claim is open.

• MGA: Managing general agent; a firm that underwrites and issues policies on a carrier's behalf.

• FNOL: First notice of loss; the intake step when a claim is reported.

• Reserve: The amount the insurer sets aside for a claim that is not yet fully paid.

• Loss ratio / loss-and-LAE ratio: Losses (and loss-adjustment expenses) divided by premium; III's 2023 P/C table is the industry baseline used above.

• AIS Program: The written AI-systems program the NAIC model bulletin expects insurers to keep for governance and consumer-harm controls.

• Federated context graph: Federato's name for the shared data structure that keeps claims, policy, and underwriting in one architecture.

• Unfair claims settlement practices: State-law standards, referenced in the NAIC bulletin, that still apply when AI supports a claims decision.

Federato Claims is a carrier-core loop, not a small-business app. If you already move documents and decisions through US Tech Automations, the useful next step is to map which of your files would have to be structured for a carrier to consume them, then put that map on an agentic workflow instead of a rebuilt stack.