Editorial Policy

This Editorial Policy describes how 2026-06-v1 — the Orbis Brief Standard — governs the production of Orbis Signal intelligence briefings across politics, finance, and technology. Every published edition is produced under this standard and records the policy version in its production metadata, so any briefing can be traced to the editorial rules in force at the time it was created.

Orbis Signal is not a wire service or headline aggregator. We publish structured analytical briefings for readers who already know the headline and need the strategic implication, calibrated uncertainty, and decision relevance behind it.

Our briefings are written for high-net-worth individuals, family offices, institutional allocators, and senior professionals who:

• Already follow major developments in their domains of interest
• Want analytical depth and second-order effects, not event recaps
• Distrust sensationalism and expect explicit treatment of uncertainty
• Make decisions — on capital, geography, reputation, or security — where framing quality matters

The editorial test is simple: does this briefing help a sophisticated reader make a better-informed decision? If it is vague, superficial, over-stated, or merely repeats mainstream narrative without adding analytical value, it should not be published.

What our voice is: calm, restrained, and analytical; comfortable with probabilistic language; structurally clear with strong hierarchy and dense signal — every sentence should earn its place.

What our voice is not: hysterical, urgent, or dramatic; hedged into uselessness; repetitive or padded; presumptuous about what the reader already knows.

We write declaratively where evidence permits, quantitatively where data exists, and forward-looking without crossing into unsupported speculation. We are neither alarming nor dismissive.

Every key judgment in an Orbis briefing carries an explicit epistemic type and confidence band. This is our most important editorial standard.

• Confirmed — directly sourced and verifiable (e.g. official releases, named institutional data). Requires a direct source citation.
• Inference— logical deduction from confirmed evidence, not stated by any single source. Signalled with language such as “suggests” or “is consistent with.”
• Probability— assessment of likelihood with explicit uncertainty (e.g. “likely,” “our assessment is”).
• Speculation — plausible but without firm evidential basis. Must be labelled as such.

Confidence bands — high, moderate, low, or insufficient evidence — reflect source quality, corroboration, and contradictions. When evidence is insufficient, we record the gap rather than speculate.

Each intelligence briefing follows a fixed section order designed for rapid extraction of analytical value:

• Executive summary — bottom line up front; the single most important conclusion, not a table of contents
• Key judgments — numbered analytical outputs with epistemic labels
• Why this matters — consequences and trajectories for the target reader
• Strategic implications — second-order effects
• Uncertainty register — what is genuinely unknown
• Decision relevance — mapped to capital, reputation, regulatory, geographic, and security dimensions where material
• Consensus gap — what mainstream coverage misses or misframes
• Watchlist — specific, actionable indicators to monitor
• Sources — tier-labelled attribution list

The following are always wrong and are flagged by our quality controls:

• Sensationalist phrases (e.g. “imminent conflict,” “game-changer,” unqualified “unprecedented crisis”)
• AI disclaimer language that breaks the editorial voice
• Empty hedging that adds no information (e.g. “remains to be seen,” “time will tell”)
• Undifferentiated predictions without analytical basis
• Padding phrases that do not carry signal
• Confirmed claims without supporting evidence or tier-1 sourcing

Each edition begins with an editorial frame that sets priority domains and guiding questions for the day. Within each topic, research plans select three to four investigation threads based on significance scoring. Threads below the significance threshold are excluded, and those exclusions are recorded as part of the production audit trail.

We prioritise developments with strategic consequence over volume of coverage. Not every headline warrants inclusion in a briefing.

When this policy changes, we increment the version string (currently 2026-06-v1). Updated policies apply to editions produced after the change; previously published editions retain their original policy version in metadata.

Related disclosures: Fact-Checking & Verification, AI Usage Disclosure.