Respond to a publication ethics crisis

---
name: respond-to-a-publication-ethics-crisis
category: write
description: Respond to a publication ethics crisis through harm containment, evidence preservation, independent claim and source review, conflict disclosure, correction or retraction, and accountable reform. Use when serious allegations challenge the truthfulness, independence, consent, provenance, or safety of published work.
---

# respond-to-a-publication-ethics-crisis

Protect affected people and the record before defending the institution.

## When to use

- Use for alleged fabrication, plagiarism, manipulated media, source abuse, undisclosed conflicts, false attribution, unsafe identification, or material data error.
- Activate qualified editorial, research-integrity, safeguarding, legal, privacy, security, rights, communications, and independent governance support.

## Preconditions

- Establish an unconflicted incident lead and independent editorial or governing authority.
- Define affected works, claims, people, sources, regions, formats, derivatives, risks, deadlines, and stop authority.
- Preserve published versions, source material, edits, communications, access, distribution, and evidence before destructive changes where safe.

## Procedure

Complete **immediate harm and evidence control**, **claim and source audit**, **independence and conflict review**, and **correction, retraction, and recovery** as one process.

1. Open a **publication ethics crisis timeline** recording works, releases, claims, sources, edits, evidence, complaints, harms, distribution, decisions, communications, and timestamps.
2. Establish protected evidence custody and source compartmentation: an independent custodian captures immutable originals, hashes and timestamps them, logs every access, revokes accused or conflicted access, separates source-identity keys, and gives reviewers redacted or pseudonymous working copies by default.
3. Require a documented two-person decision by the source-safety lead and independent review authority before unmasking a source; record reviewer funding, terms, conflicts, evidence access, publication independence, and appeal route, and contact vulnerable sources only through a verified safe channel.
4. Protect affected people through promotion stops, access restriction, identity safeguards, support, safe contact, or temporary editor's notice without assuming the allegation is proven.
5. Preserve manuscripts, notes, recordings, images, data, code, references, consent, rights, translations, models or tools, prompts where relevant, edits, approvals, and access history.
6. Establish independent reviewers and manage author, editor, funder, advertiser, source, institution, and governing conflicts plus recusals.
7. Build a **claim, source, version, and derivative reconciliation** for every material statement, quote, identity, image, datum, translation, citation, and conclusion.
8. Verify source authenticity, provenance, custody, consent, promises, corroboration, representation, quotation, context, and contradictions.
9. Audit data collection, analysis, code, charts, statistical claims, image or audio edits, synthetic or AI-assisted material, plagiarism, citation, and reproducibility.
10. Trace web, print, feed, newsletter, social, partner, translation, caption, transcript, archive, search, download, and syndicated derivatives.
11. Classify each finding as supported, imprecise, incomplete, misleading, erroneous, fabricated, plagiarized, unsafe, conflicted, or unresolved.
12. Decide no change, clarification, correction, editor's note, expression of concern, recall, retraction, apology, or investigation extension with explicit authority.
13. Publish visible versioned action, propagate it to derivatives and partners, protect people, and avoid repeating sensitive harm.
14. Notify affected people, sources, partners, institutions, regulators, or funders as required from verified facts.
15. Review contributor accountability, commissioning, source practice, fact checking, data, AI and media provenance, conflicts, approvals, corrections, and governance.
16. Monitor residual copies, harm, appeals, threats, public claims, and remediation effectiveness through closeout.

## Failure plan

- If evidence may be actively destroyed, preserve access and custody immediately through authorized independent control.
- If imminent harm continues, restrict the affected material while preserving a transparent notice and review record.
- If independent review cannot be established internally, use qualified external authority.
- If a conclusion remains unresolved, state uncertainty and keep the investigation open rather than forcing exoneration or blame.

## Worked example

An investigative series is accused of fabricated quotes, unsafe identification of vulnerable sources, undisclosed funder and author conflicts, manipulated images, unmarked AI-generated material, incorrect data analysis, and misleading translations after the work was syndicated, excerpted, downloaded, and cited in several countries. The publisher stops promotion, protects sources, preserves evidence, appoints independent reviewers, audits every claim and derivative, publishes proportionate corrections or retractions, and reforms commissioning plus provenance controls.

## Done

- A publication ethics crisis timeline verifies works, claims, sources, versions, immutable custody, compartmented identities, reviewer independence, evidence access, harms, distribution, conflicts, decisions, and communications
- A claim, source, version, and derivative reconciliation proves quotations, identities, media, data, translations, citations, consent, rights, conclusions, and unresolved findings
- A correction, accountability, and reform report demonstrates independent review, proportionate publication action, propagation, affected-person protection, notifications, appeals, governance remediation, and monitoring