Pennsylvania AG settled with GEICO over AI underwriting tied to improper policy cancellations

Pennsylvania Attorney General Dave Sunday announced a settlement with GEICO on May 22, 2026, after an investigation found the insurer's AI tool for selecting new policyholders for underwriting review caused customer confusion and unfair policy cancellations. The AI selected a policyholder for review who submitted documents she believed were adequate, but GEICO failed to inform her the submission was insufficient and cancelled her policy without adequate notice, leaving her unknowingly driving uninsured. GEICO agreed to extend document submission deadlines, reduce verification requirements, and align with state AI guidance without admitting any violation of law.

GEICO · Incident May 22, 2026 · Indexed Jun 4, 2026 · 3 sources

The AI tool selected a new policyholder for review, triggered a documentation request, then the automated process cancelled her policy without ever telling her the documents she submitted were insufficient.
What
Pennsylvania Attorney General Dave Sunday announced a settlement with GEICO on May 22, 2026, after an investigation found the insurer's AI tool for selecting new policyholders for underwriting review caused customer confusion and unfair policy cancellations.
Incident date
May 22, 2026
Who
GEICO
Failure mode
Agentic Action Error
AI surface
Agentic Workflow
Severity
Medium

What happened

GEICO used an AI tool with artificial intelligence features to select new policyholders for further underwriting review during the standard 60-day policy review period. In the case that triggered the investigation, the AI selected a West Philadelphia policyholder who was then required to submit additional documentation or face cancellation. The policyholder submitted what she believed were the correct documents, but GEICO did not inform her that the submission was inadequate. Her policy was cancelled without adequate notice, causing her to unknowingly drive uninsured. Pennsylvania AG Dave Sunday announced the settlement on May 22, 2026, requiring GEICO to modify its AI-driven cancellation process and comply with state AI guidance.

What broke inside the model

Failure path · mode profile · Agentic Action Error
  1. 01 · TriggerAn agent plans a multi-step task.
  2. 02 · Model stepIt chooses a wrong or destructive action.
  3. 03 · Control gapNo confirmation gate guards the write.
  4. 04 · FailureThe action commits to a system of record.
  5. 05 · ConsequenceData is changed or destroyed irreversibly.

A wrong action commits, and the step is written before anything can stop it.

The AI-driven workflow selected a customer for review and triggered a documentation request, but the downstream process lacked adequate feedback mechanisms to inform the customer when submitted documents were insufficient. The system proceeded to policy cancellation without ensuring the customer understood the deficiency, effectively treating the lack of adequate documentation as tacit consent for termination. The process was especially confusing for lower-income new policyholders who found the dual verification requirements difficult to navigate.

Public visibilityMedium
Regulatory exposureActive
Customer impactClass-wide
Financial impactUnknown
Time to disclosureMonths
  1. PrimaryAG Sunday and GEICO Agree on Improvements to Remedy Customer Confusion and Prevent Unfair Auto Insurance Cancellationsattorneygeneral.gov
  2. PressAfter Complaint, GEICO Agrees to Modify Cancellation Process That Uses AIinsurancejournal.com
  3. PressGEICO settles suit over AI-driven auto insurance cancellationsfox43.com
Permalinkhttps://failureindex.ai/failures/pennsylvania-ag-settled-geico-ai-underwriting
CitationAI Failure Index. "Pennsylvania AG settled with GEICO over AI underwriting tied to improper policy cancellations" (FI-0118). Realm Labs. https://failureindex.ai/failures/pennsylvania-ag-settled-geico-ai-underwriting (indexed Jun 4, 2026).
Share cardA branded image of this record for posts and slides.

Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0118. Full dataset at /data.

Note from Realm Labs, the Index steward

How Realm would have caught this

Controls for this failure mode
  • Prism
  • OmniGuard
  • AgentRealm

Realm can sit inline on the agent's action path and require that a destructive or high-consequence action clears a real check before it executes, so 'delete and recreate' or a wrong write is stopped at the moment of intent, not explained in the post-mortem.