Editorial Standards
Stop Flock Cams Now · Last updated: 2026-08-23 · Licensed CC BY 4.0
Our name states a position. Our data does not. This document is how we keep those two things separate and both credible — the mission-named, fact-credible model. It governs everything SFCN publishes; sourcing rules live in source-standards.md, error handling in corrections-policy.md.
1. Two layers, visibly separated
The data layer — the map, camera dossiers, jurisdiction statistics, rankings, dumps, and the API — is neutral. It reports documented facts with provenance and confidence tiers (methodology.md), uses descriptive language ("documented," never "infested"), links its methodology from every page, and is written so that a police chief, a privacy activist, and a federal judge can all cite it without embarrassment.
The advocacy layer — "Take Action" sections, opinion posts, campaign explainers — is clearly labeled as advocacy, visually and structurally separated from data pages, and limited to lawful civic participation: contacting officials, attending and speaking at public meetings, public-records requests, petitions, elections, and organizing. Advocacy content never appears inline in the data layer; data pages may link to a Take Action section, marked as such.
Advocacy content is held to the same factual standards as everything else. We argue from accurate premises or not at all. And in both layers, we never encourage, tolerate, or wink at vandalism, tampering, interference, trespass, harassment, or any illegal act (terms-of-service.md §4).
2. Claim labeling: ALLEGATION / FINDING / COURT DECISION
Any claim of wrongdoing, misuse, illegality, or misconduct — about an agency, an official, an officer, a company (including Flock Safety), or anyone else — carries exactly one of these labels wherever it appears (data fields, incident tracker entries, articles, share cards):
| Label | Meaning | Standard |
|---|---|---|
| ALLEGATION | A claim asserted by someone — a plaintiff, a complainant, a news source — that has not been established by an authoritative process | Must cite who alleges it and where; must meet the two-source/adverse-claim rule in source-standards.md; phrased as attribution ("the complaint alleges…"), never as fact |
| FINDING | A determination by an authoritative non-judicial process: an audit, inspector general, internal-affairs conclusion, regulator, or our own documented analysis of primary records | Must cite the finding document; our own findings must show their full workings and data |
| COURT DECISION | A ruling by a court | Must cite the case, court, and date; must state procedural posture accurately (trial ruling vs. appeal pending vs. final; e.g., a district-court ruling under appeal is described as exactly that) and be updated when the posture changes |
Related discipline: settlements are settlements (not findings of liability); charges are allegations; "accused of" is never shortened to "did." Mislabeled claims are treated as critical corrections (corrections-policy.md §5).
3. AI-assisted content
- Any figure, identification, or text produced with AI assistance is labeled "AI ESTIMATE" at the point of display until a human has verified it against sources — and AI camera-photo identification is always labeled AI ESTIMATE and never auto-verifies a record.
- Nothing AI-generated publishes without human review. The research pipeline routes all AI extractions through a human approval gate (../docs/AI_RESEARCH.md); AI output is a lead, never a source (source-standards.md).
- Human-reviewed, source-verified facts that originated from AI extraction cite the underlying source, and the record notes AI-assisted extraction in provenance.
4. Fairness and subjects of coverage
- Subjects of adverse editorial coverage get a documented opportunity to comment before publication where feasible, and their responses are included or summarized fairly.
- We cover this beat's facts wherever they land — including facts favorable to Flock Safety or ALPR-using agencies (e.g., policy reforms, cleared allegations). Selective omission is an accuracy failure.
- Individuals are covered only in their official capacities and only as documented in sources; private individuals are not named absent overwhelming public interest, and doxxing-adjacent detail is never published (community-guidelines.md).
- We identify ourselves honestly when gathering information. No pretexting.
5. Language and presentation rules
- Absence caveat: any statistic that could imply completeness carries the "absence of a marker ≠ absence of cameras" caveat, verbatim or linked (disclaimer.md).
- Rankings say "most-documented," with methodology linked, per methodology.md §5.
- Numbers carry dates and sources; superlatives carry evidence or don't appear.
- "Flock" is used nominatively to refer to Flock Safety, Inc. and its products; we do not use their logo or trade dress, and the non-affiliation disclaimer appears sitewide.
- Headlines must be supportable by the body. Share cards are held to the same standard as pages — the card footer carries source and date.
6. Conflicts and independence
SFCN accepts no funding from surveillance vendors or their agents, discloses any funder with an interest in coverage, and keeps editorial decisions independent of donors. Contributors with a conflict (e.g., employment by an agency or vendor covered) must disclose it on affected submissions.