Triggers, exclusions, and the genuine-dispute line.

Bad faith turns on whether the carrier owed the claim, and that begins inside the policy. JustineAI™ is being designed so the reasoning core reads the full policy form — insuring agreement, conditions, exclusions, endorsements — against the loss facts, and tests the carrier’s coverage position the way a court would. A coverage facet will map each trigger and exclusion the carrier invoked to the language that supports or defeats it, then probe the genuine-dispute defense: was the carrier’s denial a reasonable reading, or a constructed dispute the file itself contradicts?

What IB is designed to do.

1. 01

   The reasoning core is being built to read the entire policy as one instrument — declarations, insuring agreement, conditions, exclusions, every endorsement — rather than the single clause a denial letter chooses to quote.

2. 02

   A coverage facet will trace each exclusion or condition the carrier invoked back to its operative language and to the loss facts, surfacing where the carrier’s reading strains the words on the page.

3. 03

   The genuine-dispute doctrine will be tested structurally: the facet weighs whether a reasonable adjuster could hold the carrier position, or whether the claim file shows the dispute was manufactured after the decision to deny.

4. 04

   Ambiguity and reasonable-expectations arguments are designed to surface with the competing constructions laid side by side, each anchored to CourtListener-grounded authority for the controlling jurisdiction.

5. 05

   The interpretation is the attorney to adopt or reject; the platform proposes the coverage theory and its citations, and the attorney attests before any of it informs a demand.

The AI reasons; the attorney decides.