Truth Shield Advanced — Candor Trust & Safety

A real newsroom-tier deep analysis, live in the portal: an unverifiable "leaked email" claim about an airline's overbooking practices scores 15% (Likely False) — with Sonnet citing real DOT overbooking data, naming the evidentiary weaknesses, and giving the journalist concrete next steps before publishing.

What it adds

Truth Shield Advanced is the journalist and newsroom tier, built for cases where a surface-level score isn’t enough. On top of the base tier’s single-pass check, it runs a deeper, Sonnet-powered investigation that returns per-claim citations with their own assessment and basis, a source credibility analysis with named red flags, an influence-campaign and coordinated-amplification check, a timeline/virality assessment, and a set of concrete next actions a journalist can actually act on — including a direct publishability recommendation, not just a score.

Verified behaviors

Every claim below maps to a dated behavioral proof against the live production service.

The journalist-tier deep analysis provides Sonnet-powered investigation with per-claim citations, source credibility assessment, influence-campaign detection, and specific newsroom recommendations — not just a single score.
method: a fabricated health-misinformation claim was submitted to the live deep-analysis endpoint using a real journalist-tier API key · observed: returned a 3-entry citations array with a specific assessment and basis for each sub-claim, a source-credibility analysis with 8 named red flags, an influence-campaign analysis (detected, with likely origin and coordination signals), a timeline/virality assessment, and 6 concrete journalist recommendations including a named contact to verify with and a "do not publish yet" recommendation with its own reasoning · VERIFIED July 28, 2026

Honest limitations

Deep analysis takes materially longer than the base tier — this is deliberate depth, not a fast check, and should be used for stories that warrant the wait.
A "do not publish yet" or "likely false" assessment is an investigative starting point for a journalist’s own verification, not a replacement for it.

Download the white paper (PDF)

Evaluate it against your platform

Book a demo

Where this fits

Candor works at the intersection of gaming safety, age assurance, privacy, parental consent, cryptographic evidence, and auditable trust & safety decisions. Related:

Truth Shield — credibility analysis with the reasoning shown
Never a black-box score; open-source core.
Tips — a privacy-first secure channel for journalists and sources
Source protection by architecture, not policy.
The Candor Architecture — screen-before-store with auditable trust and safety decisions
The embedded gateway and full gate chain, end to end.