AI Assistance & Source Attribution

Editorial AI & Source Attribution Policy

Document key: ai_attribution

Version: 1.0

Effective date: July 27, 2026

1. Why this policy exists

Some EddieRod.com editorial work is written with AI assistance. Readers have a right to know when AI was involved, how third-party reporting is handled, what may publish without human review, and who remains accountable.

2. Who may use editorial AI

Only authorized administrators, moderators, and editors may activate editorial AI features. Ordinary members cannot send their private drafts, Bulletins, riffs, comments, or profile information to OpenAI through the editorial workflow.

Authorized editorial users may send third-party source titles, links, excerpts, images or image metadata, and writing instructions to OpenAI to prepare or evaluate an article draft.

EddieRod.com does not use member User Content to train AI models without separate express authorization.

3. How the editorial pipeline works

We monitor selected public sources, including news outlets, blogs, and feeds. When relevant material appears, the system captures limited source information or excerpts called “clippings.” An AI writer may prepare a draft using those supplied sources as its factual basis.

Our default posture is strict attribution. The model is instructed not to invent facts, quotations, names, numbers, dates, or claims and to attribute material claims to supplied sources. Thin, conflicting, or uncertain sourcing should be identified rather than hidden.

4. Phase 4 automated publishing

At Phase 4, the automated system may:

• evaluate real editorial jobs against a versioned publication policy;

• publish an eligible article to a public Bulletin page without a person reading it first;

• make the article eligible for the EddieRod.com Feed and RSS; and

• record the decision inputs, policy version, outcome, and rollback information.

Automated publishing does not send email. Email distribution remains outside the automated publisher because an email cannot be recalled in the same way as a webpage or Feed item.

Anything that fails a blocking policy check remains unpublished or goes to a human. Automated publication can be stopped through an emergency control and an article can be withdrawn or rolled back.

5. Labels and disclosure

An article drafted with AI assistance carries a visible AI-assisted label on the article page and in the Feed.

An article published automatically without human review carries an additional visible notice stating that it was automatically published and was not reviewed by a person before publication.

A later human review or correction does not erase the historical fact that an article was originally published automatically.

6. How we treat sources

Our rules are:

• Attribute: identify the publisher, author, or source of material reporting where reasonably available.

• Link: link to the original public source when possible.

• Use only what is needed: quote or summarize the minimum reasonably necessary to report, comment on, compare, or explain the subject.

• Add independent value: do not substitute for the source or merely rewrite it sentence by sentence.

• Do not conceal dependence: do not paraphrase around an attribution to make another publisher’s reporting appear original.

• Respect exclusion requests: evaluate and honor a publisher’s reasonable request to stop ingestion or use where appropriate.

• Escalate uncertainty: avoid automatic publication when source ownership, accuracy, quotation, privacy, defamation, or fair-use risk is unclear.

There is no universal safe word count or percentage. The amount used depends on purpose, nature, necessity, market effect, and other circumstances. Editorial settings therefore use minimality, attribution, transformation, and source-replacement checks rather than a claimed numeric safe harbor.

7. What AI does not do here

• It does not conduct original interviews or make first-hand observations.

• It does not decide EddieRod.com’s editorial positions.

• It does not send automated email.

• It does not independently moderate member accounts or community content.

• It does not make decisions about employment, housing, credit, health care, education, insurance, criminal justice, or basic necessities.

8. Accountability

EddieRod.com is responsible for material published under its name, including work published automatically. Automation is a tool chosen by the operator and is not treated as a defense.

Editorial responsibility rests with the Editorial Director.

9. Corrections, withdrawal, and removal

Send a correction, attribution, rights, or editorial complaint to focus@eddierod.com and identify the article and disputed statement.

Our usual response is:

• Factual error: correct the article and add a dated visible correction note describing the material change.

• Attribution error: correct the source or quotation and add a note when the error was material.

• Serious, disputed, or unsafe article: withdraw or remove the article while reviewing it, then republish with a correction or leave it withdrawn or removed with an appropriate explanation.

• Copyright complaint: follow the DMCA & Copyright Policy or another applicable rights process.

The decision to correct, withdraw, or remove depends on severity, confidence, safety, legal risk, and whether a corrected version can be reliable. When an automated article fails, we also review whether publication policy or source controls should change.

10. Source transparency

Each Bulletin should identify material sources through links, a source trail, or another reader-accessible method, except when a source must lawfully or ethically remain confidential.

11. Changes

We publish a new version and effective date when this policy materially changes.

12. Contact

Editorial Director, corrections, and legal notices: focus@eddierod.com

Telephone: 512-387-0093

Mail: 3816 South Lamar Blvd, Building 22, Austin, Texas 78704, United States