Trust
Security & data
Last updated: July 2026 · CiteSafe Scout 1.0 Beta · Honest about what ships today
What CiteSafe is
CiteSafe is the trust layer for legal AI — independent citation verification after any AI or human draft. We do not train a general legal chatbot on your briefs.
Beta data handling
- Much of Scout 1.0 Beta scan history stays in your browser (device-local storage) unless a server feature is explicitly enabled.
- Site traffic uses standard HTTPS (TLS) via our host (GitHub Pages / CDN edge).
- We do not sell your documents or personal data.
- We do not use customer briefs to train a public generative model.
What we do not claim yet
No SOC 2 Type II badge on this page until audited. No enterprise SSO, DPA pack, or on-prem deployment in the public beta. Those track with Clerk / Bench and firm customers — not as marketing fiction.
Sources used for checks
Multi-layer stack: fabrication heuristics, CiteSafe authority corpus, CourtListener (Free Law Project), Caselaw Access Project (Harvard), and GovInfo / public statute signals when the API is hosted. Offline Scout uses corpus + heuristics only and says so in the UI. We are not a full commercial legal research library, and we do not pretend to be. Existence ≠ good law or holding support.
Human in the loop
Statuses are Verified, Needs Review, and Could Not Verify — designed so attorneys still review before filing. We never market “hallucination-free” guarantees.
Contact
Security or data questions: security@citesafe.co