A suit alleges State Farm's fraud-detection AI disproportionately flagged Black homeowners' claims
In Huskey v. State Farm Fire and Casualty Co., filed December 14, 2022, two Black homeowners alleged that State Farm's machine-learning fraud-detection algorithms assigned higher risk scores to Black policyholders using race-correlated proxy inputs, routing their claims into heightened scrutiny and causing significant delays. The complaint cites evidence that Black policyholders were 39 percent more likely to submit extra paperwork, while white homeowners were nearly a third more likely to have claims processed within a month. The court denied State Farm's motion to dismiss the disparate impact claims in September 2023, and discovery remains ongoing.
Race-correlated proxy variables fed into a fraud-scoring algorithm turned historical housing bias into a self-fulfilling loop that flagged Black homeowners for extra scrutiny while fast-tracking similar white claims.
Key facts
- What
- In Huskey v.
- Incident date
- Dec 14, 2022
- Who
- State Farm
- Failure mode
- Policy Violation
- AI surface
- Agentic Workflow
- Severity
- High
What happened
Jacqueline Huskey and Riian Wynn filed a putative class action on December 14, 2022, alleging that State Farm's algorithmic claims-processing and fraud-prediction systems discriminated against Black homeowners in the Midwest. Huskey experienced a two-month delay in benefit payments that caused additional water damage to her kitchen and bathrooms, while Wynn's roof claim took approximately three months longer to process than her white neighbor's identical claim, forcing her to move out of her home. A 2021 YouGov survey cited in the complaint found that Black policyholders were 39 percent more likely to have to submit extra paperwork and white homeowners were nearly a third more likely to have claims processed within a month. The court denied State Farm's motion to dismiss the core disparate impact claims in September 2023, and discovery remains ongoing.
What broke inside the model
- 01 · TriggerA prompt pushes against a deployment boundary.
- 02 · Model stepThe model produces the disallowed output.
- 03 · Control gapNo enforcement blocks it at generation time.
- 04 · FailureThe output crosses the policy line.
- 05 · ConsequenceA limit the business set is breached in public.
The output crosses a policy boundary the deployment had defined.
State Farm's FRISS fraud-detection software, deployed through Duck Creek Technologies, assigned policyholders a risk score using inputs that functioned as proxies for race, including biometric data, geolocation, social media presence, browser history, and historical housing data already infected with racial bias. These biased inputs caused the algorithm to disproportionately classify Black homeowners as high-touch claims requiring additional scrutiny, while white homeowners with similar claims were more often classified as low-touch and paid out promptly. The system lacked sufficient oversight or bias auditing to detect and correct this disparate impact.
What it cost
Sources
- PrimaryState Farm Algorithm Bias Lawsuit | Sanford Heisler Sharp McKnight, LLPsanfordheisler.com
- PressState Farm Must Face Race Discrimination Suit Over Algorithmsnews.bloomberglaw.com
- Court FilingHuskey v. State Farm Fire and Casualty Company - 1:22-cv-07014classaction.org
Cite this entry
https://failureindex.ai/failures/suit-alleges-state-farm-fraud-detectionAI Failure Index. "A suit alleges State Farm's fraud-detection AI disproportionately flagged Black homeowners' claims" (FI-0113). Realm Labs. https://failureindex.ai/failures/suit-alleges-state-farm-fraud-detection (indexed Jun 4, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0113. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
Realm compares what the model is about to output or do against the policy that governs the deployment, in real time, and can deny or redact the action before it takes effect, which is the gap an after-the-fact review never closes in time.