The trust layer
D4NN is built around a simple commitment: every signal should be inspectable. Status, confidence, and relevance are computed separately — and visibly.
Status
What kind of claim is this? Confirmed, likely, developing, disputed, false, context.
Confidence
How much should you trust this signal, based on source quality and corroboration?
Relevance
How much does this matter to you, based on your role, topics, and watchlists?
Reality is not a binary. We use eight statuses so you can act with the precision the situation deserves.
Confirmed
Independently verified by multiple primary or high-trust sources.
Likely
Strong corroborating evidence; not yet primary-source confirmed.
Developing
Active situation; facts shifting in real time.
Disputed
Credible sources disagree. Both framings shown.
Unverified
Claim circulating but no reliable evidence yet.
Misleading
Technically true but framed in a way that distorts.
False
Disproven by primary evidence or competent authorities.
Context
Background or framing — not a claim about the present.
Every source is rated by tier. Confidence rises as evidence climbs the trail — and is never higher than its weakest required link.
Primary
Direct documents, filings, official records, raw data.
Institutional
Established outlets, regulators, peer-reviewed research.
Expert
Domain practitioners with track records and accountable handles.
Social
Open web and platforms. Weighted lower; never sufficient alone.
Every signal in your briefing carries an explicit reason it was surfaced to you — the topic, watchlist, or signal-of-signals that matched. Personalization without opacity.