AI Disclosure

Last updated July 15, 2026

Where AI is used

AI may classify a preliminary policy category, identify evidence gaps, structure facts, draft filing language, and validate generated output.

What is not shown

The product displays useful analysis stages and conclusions, not private model chain-of-thought. Locked advanced stages show what additional work a paid product performs without pretending that skipped analysis occurred.

Human responsibility

You control the evidence set, can edit generated output, and must verify the final filing. AI does not make the platform's decision or replace professional legal advice.

Quality safeguards

Prompts, policies, rubrics, model selections, outputs, and evaluations are versioned. The system is designed to reject fabricated evidence, unsupported citations, and removal guarantees.