260% spike in NAICS 56 injury rate vs national mean.
1,848 NAICS-state-year cells from BLS SOII public reference data, run through our WC integrity pack: NAICS injury-rate outliers, state DAFW concentration, year-over-year anomalies. 207 deterministic findings - including a NAICS 56 (Administrative and Support / Waste Management) injury rate running 260% above the national mean. Real federal reference data; T1 baseline available to every WC carrier on day one.
JIL is not a claims management system, not a TPA, not a medical bill review service, not a recovery vendor. It is a payment-integrity layer covering both the premium side (employer underpayment via misclassification) and the claim side (claimant, provider, and attorney fraud). Premium fraud is a category most fraud-detection products do not address; we surface it from the same engine that scores claims.
Here's what we surface for you.
If you're a workers' compensation insurance carrier, an SIU lead, an underwriting actuary, a state DOI fraud bureau, or an employer payroll auditor, this is what JIL detects in workers' comp flows. JIL plugs in at three places (premium underwriting, claim adjudication, recovery operations), runs 500+ checks (148 attestation catalog + 27 WC-specific), and produces sealed evidence per finding. Pulls live data from BLS occupational injury statistics + 8 federal datasets.
Premium evasion (employee misclassification)
Employer claims they have 12 employees in their WC application. Their actual federal payroll filings show 47. Or they classify workers as 1099 contractors who legally meet the W-2 employee test. JIL cross-checks against state employment tax records, federal W-2 filings, and prior-year payroll signals to surface premium evasion before policy renewal.
Ghost employees on the claim
Employer files a WC claim for an employee whose Social Security record shows no actual employment with this employer in the relevant period. Or the employee's address pattern doesn't match the work site. JIL surfaces ghost-employee patterns before the indemnity payment releases.
Doctor / clinic kickback rings
Same medical provider treats 47 WC claimants in 6 months, all with similar diagnosis patterns, all referred by the same attorney, all filing claims against employers in the same industry. JIL surfaces the kickback ring with the entity-graph evidence - the kind of pattern your SIU finds in years of manual work.
Modified-duty fraud / malingering pattern
Claimant on modified duty / total temporary disability for an injury whose typical recovery is 6-12 weeks, but the claim has been open 18 months. JIL cross-checks against BLS occupational-injury median recovery times for the specific injury + occupation + age cohort and flags the outlier.
Subrogation recovery support
WC claim resulted from a third-party tort (auto accident with another driver, premises liability at a non-employer location). JIL surfaces subrogation-recovery candidates by cross-checking the claim circumstances against vehicle records, premises ownership records, and prior litigation.
Sealed evidence per finding
Every check produces a sealed evidence record. Useful for SIU investigation, premium-audit defense, IME challenges, and (if it escalates) civil litigation or criminal referral. State DOI fraud-bureau referral-ready.
How to use it: integrate at three points - underwriting (premium-evasion check at quote / renewal), claim adjudication (per-claim screening), and recovery (subrogation candidate identification). Per-claim + per-policy pricing, with volume tiers. See the live POC with 207 findings against synthetic WC claim batches.
260% spike in NAICS 56 injury rate vs national mean.
On day one, wc-engine surfaced a 2.6x national-mean spike in NAICS 56 (Administrative Support) injury rate in 2023, paired with a 60% drop in NAICS 49 (Warehousing) in 2022 that smells like underreporting. 207 Tier 1 findings: 124 NAICS-outliers, 83 state-concentration patterns (California holds 65% of NAICS 62 health-care DAFW cases; Texas holds 55% of NAICS 72 accommodation cases). All from BLS-shaped public injury rates.
What we uncovered (top 5)
| Subject | Category | Finding |
|---|---|---|
| NAICS 56 (2023) | Admin support: 6.76 per 100 FTE | 2.6x national mean - high severity |
| NAICS 49 (2022) | Warehousing dropped 60% YoY | Suspected underreporting |
| California in NAICS 62 | 65% of national health-care DAFW | State concentration |
| Texas in NAICS 72 | 55% of accommodation DAFW | State concentration |
| 124 + 83 + (year-over-year) | Three distinct check categories | Cross-cutting WC signal |
BLS Survey of Occupational Injuries (SOII)
Annual BLS SOII reference rates by NAICS, year, and state. No subscription, no DUA, no per-record licensing. The full ingest manifest is replayable bit-identically.
Deterministic pass
Tier 1 ran a single deterministic pass against the ingested public dataset. SQL aggregates only - no stochastic LLM in the verdict path. Ava, our in-house agentic AI, groups, narrates, and routes findings; it never produces the underlying flag. Same kernel that ships the other 7 verticals.
What this means for your business
For a workers' comp carrier's actuary, NCCI experience-mod analyst, or state DOI underreporting investigator, this is the NAICS-state-year anomaly queue your existing rate-making process cannot produce because the cross-state YoY comparison requires the consolidated BLS series. Each finding ships with a sealed CREB™ suitable for NCCI submission, OSHA inspection-priority referral, or rate-filing supporting evidence.
What Tier 2 unlocks. Customer engagement adds carrier-supplied claim event records, employer payroll records, named-medical-provider data. Tier 2 decomposes the YoY anomaly into volume vs frequency components, identifies the named insureds driving each cohort, and produces a sealed CREB™ suitable for experience-mod adjustment or OSHA referral.
Workers' Compensation Integrity at a glance.
Where the integrity layer sits, what it produces, and how the sealed CREB™ flows back to the buyer's existing systems.
Where this product earns its place.
The strategic case for Workers' Comp as a JIL line of business - what makes the wedge defensible, what makes it economically meaningful, and how it compounds with the rest of the platform.
~90% architecture transfers
Healthcare provider-fraud detection (chiropractic mills, opioid prescribers, IME shopping) transfers nearly intact. P&C claim adjudication and SIU evidence framework transfers nearly intact. H1B classification logic informs premium fraud.
Buyer base parallels Medicaid agencies
Texas Mutual, California SCIF, NYSIF, Pinnacol, Ohio BWC have procurement structures comparable to state Medicaid agencies. The 50-state expansion playbook from healthcare maps directly.
Federal compression via OWCP
DOL Office of Workers' Compensation Programs administers four federal WC programs (FECA, EEOICPA, BLBA, LHWCA). Federal partner channel (Accenture / Leidos / Peraton) compresses procurement timeline from 18-24 months to 60-120 days.
Underserved category leadership
Most WC fraud detection focuses on the claim side. Premium fraud (employer misclassifying workers, underreporting payroll, misrepresenting business operations) is comparable in dollar magnitude (25%-40% of total) but underserved by tooling. JIL leads the category.
Net-new checks, sealed evidence.
A representative slice of the Workers' Comp-specific check pack. Each one runs in the same five-stage pipeline as the rest of the platform - intake, profile load, parallel checks, verdict, sealed CREB™ - and ships with a 14-of-20 BFT signature, a CourtChain™ L1 anchor, and a reproducibility manifest pinning the exact check-logic version.
NCCI Class Code Misclassification
Cross-references declared NCCI class code against actual operations - job descriptions, payroll, OSHA logs, NAICS code. Identifies higher-rated classes. NCCI Scopes Manual lookup.
Payroll Reconciliation
Reconciles premium-audit-declared payroll against W-2, IRS 941, and state UI filings. Flags underreporting. Accounts for legitimate exclusions (overtime premium portion, certain bonuses).
Independent Contractor Misclassification
Applies state-specific worker-classification test (ABC test in CA/MA/NJ, common-law test elsewhere). Flags 1099 relationships indicative of employee status.
Experience Modification Manipulation
Detects entity-name-change patterns indicative of experience-mod evasion. Detects policy churn timed to avoid loss reporting. Cross-references owner and officer continuity.
Functional Capacity vs Disability
Cross-references claimed disability against publicly observable activity (DMV, real estate transactions, social media) inconsistent with that disability level. Asset-intelligence integration.
IME Selection Pattern
Detects IME providers with statistically anomalous opinion patterns (95% disability-supporting when claimant-referred, 95% disability-denying when carrier-referred).
Capper and Runner Detection
Detects illegal client solicitation through third parties. Cross-references state bar disciplinary records. CA Insurance Code 1871.7 and equivalents.
OSHA-WC Reconciliation
Reconciles employer OSHA 300/300A logs against WC claim filings. Flags OSHA-recordable injuries not appearing in WC system, indicative of claim suppression.
Who runs this in production.
The buyer pattern for Workers' Comp - who carries the budget, who carries the regulatory exposure, and how the engagement starts. Most first engagements are a Retroactive Proof Audit on a defined lookback window; Pre-Settlement integration follows once the check pack is calibrated to the customer's profile.
Aggressive fraud-fighting mandates
Texas Mutual ($2B+ premium, largest US WC carrier), California SCIF, NYSIF, Pinnacol Assurance, Ohio BWC (state monopoly), Washington L&I (state monopoly), North Dakota WSI, Wyoming WCD. Aggressive fraud-fighting mandates.
Carriers and TPAs
Travelers, Hartford, Liberty Mutual, AmTrust, Berkshire Hathaway Specialty, Zurich, Chubb, AIG, Old Republic. Specialty: Employers Holdings, Amerisafe, ICW Group. TPAs: Sedgwick, Gallagher Bassett, Crawford, ESIS, Broadspire.
OWCP and OIG
DOL Office of Workers' Compensation Programs (FECA $2.5B+ annual, EEOICPA, BLBA, LHWCA). DOL OIG. VA OIG. DOE (EEOICPA). DOL MSHA. Federal partner channel via Accenture / Leidos / Peraton.
High-risk industries
Large self-insured employers in high-risk industries - construction (D.R. Horton, Lennar, Bechtel, Fluor, Turner), energy (ExxonMobil, Chevron, ConocoPhillips, Halliburton), large manufacturing, logistics (UPS, FedEx, Walmart, Amazon).
Four-SKU model. No percentage. No contingency.
Pricing carries over from the canonical four-SKU model unchanged - Retroactive Scan (flat fee), Retroactive Proof Audit (with credit-back against the next subscription tier), Pre-Settlement Subscription (tiered annual), and per-case CREB™ bundles (Tier 3 court-ready evidence). Asset Intelligence is the standard fifth SKU where the vertical needs it.
No percentage of recovery. No contingency. No success fees. JIL is detection and proof, not recovery. Recovery sits with the customer or its existing partners (subrogation counsel, recovery vendors, regulators). The structure is what allows JIL to operate as neutral integrity infrastructure across plaintiffs and defendants, payers and payees, regulators and the regulated, on the same case.
Ready to scope a Workers' Comp engagement?
Initial briefings are 60 minutes. Retroactive Proof Audit lookback windows, check-pack profile design, and integration runbook are available under NDA. We start where the buyer's procurement gate is shortest.
One kernel. Eight industries.
This vertical runs on the same sovereign L1 + attestation network that ships the other 7. Kernel age: 18+ months. Adding a vertical: ~1 week. Competitor moat: build the kernel first.
See also -- fraud-pillar counterparts
P&C and Workers Compensation Fraud Detection -- If you're a state WC fund or carrier-side SIU running fraud detection at scale, see /products/pc-insurance-fraud for the 26-check engagement model.