Policies that bind production

Ask a question

Which rule do you need?

Six documents. Start with the question; the statute is the destination.

1. 01Does a human write or check this?AI Usage Disclosure
2. 02What happens if a briefing is wrong?Corrections Policy
3. 03Who pays for this, and who can lean on coverage?Conflict of Interest Statement
4. 04Where do claims come from?Sources & Attribution Policy
5. 05How do you decide what is true without a human desk?Verification Policy
6. 06What are you allowed to say, and for whom?Editorial Policy

What governs each step

Production, bound by policy

Each edition passes through a structured, multi-stage pipeline. Web-grounded research collects and tier-labels sources; analytical models assess evidence and draft briefings; automated quality controls validate source traceability, epistemic discipline, continuity, and editorial standards before publication. These policies state the rules we follow so you can evaluate our briefings against them. The illustrated path is on Methodology.

1. Research

   Live web research extracts claims with tier-labelled source URLs, locked into an evidence record for each topic.

   Sources & Attribution Policy
2. Assess & compose

   Analytical models derive judgments from that evidence, then render structured briefings with explicit confidence and epistemic typing.

   Editorial Policy
3. Verify & publish

   Rule-based and model-assisted quality checks — including continuity for recurring storylines — run before briefings are published.

   Verification Policy
4. Automation

   There is no human author, editor, or fact-checker in the production loop between scheduled pipeline runs and publication. Standards are enforced in architecture and validation code.

   AI Usage Disclosure
5. Independence

   We do not sell advertising, native content, or editorial access. When paid plans are offered, revenue comes from reader subscriptions.

   Conflict of Interest Statement
6. Repair

   Substantive errors that affect reader understanding are investigated and disclosed. Analytical disagreement in good faith is not a correction.

   Corrections Policy

The archive

Policies & disclosures

Six documents covering editorial standards, verification, corrections, sourcing, conflicts of interest, and AI usage.

• VoiceEditorial PolicyThe Orbis Brief Standard governs every edition. Policy version is recorded in production metadata.Who we write for, how we calibrate language, and the Orbis Brief Standard that shapes each edition.Read the statute
• TruthVerification PolicyNo traditional fact-check desk. Traceability, a locked evidence record, and a fail-closed quality gate.Evidence locking, automated integrity checks, and the eight-dimension quality scorecard — not third-party fact-check APIs or manual claim-by-claim review.Read the statute
• RepairCorrections PolicySubstantive errors are investigated and disclosed. Analytical disagreement in good faith is not a correction.Reporting channels, our review process, and how corrections are applied to live and archived editions.Read the statute
• ProvenanceSources & Attribution PolicyEvery material claim traces to a public URL in the locked evidence record. Displayed URLs are a subset of that record.Source tiers, discovery preferences, evidence traceability, and how attribution appears in each briefing.Read the statute
• MoneyConflict of Interest StatementIndependent publisher. No ads, no native, no pay-for-coverage. When paid plans exist, revenue is subscriptions.Publisher independence, absence of pay-for-coverage arrangements, and disclosure obligations.Read the statute
• AutomationAI Usage DisclosureFully automated production. No human author, editor, or fact-checker between scheduled runs and publication.How AI is used in production, what humans do and do not review, and how outputs are held to editorial standards.Read the statute