Trust Center · Sourcing · Corrections
Redacted Forge treats historical storytelling as an investigation. A compelling story is not enough: material conclusions should be traceable to records, source context, counterevidence, and an explicit statement of what the evidence does—and does not—support.
Search results, competitor videos, transcripts, comments, rumors, podcasts, social posts, and repeated internet claims can point toward a question. They do not become historical proof merely because they are popular or repeated.
For material claims, Redacted Forge prefers primary records, official records, archives, scholarly work, reputable reporting, datasets, archaeology, oral histories, and affected-community sources appropriate to the question.
The narration is not allowed to be more certain than the record. Strong criticism requires stronger evidence; weak or incomplete evidence stays investigative rather than being upgraded for drama.
A Redacted File should identify the strongest good-faith alternative explanation, relevant contradictory evidence, source limitations, and uncertainty. Power or notoriety increases scrutiny; it does not create a presumption of guilt.
A document in an archive proves that the document exists in that collection; it does not automatically prove every statement inside it. A proposal is not the same as an executed operation. An allegation is not a conviction. A settlement is not automatically an admission. Custody is not endorsement.
Reconstructed scenes, generated visuals, inference, hypothetical staging, and editorial illustration must not be presented as authentic archival evidence. Redacted Forge may use AI and deterministic animation as production tools, but authentic records remain distinguishable from reconstruction.
Named historical people should not receive invented private dialogue, private thoughts, motives, or emotional certainty presented as fact.
Corrections are part of the authority system, not an embarrassment to hide. Material corrections should preserve version history, identify what changed, propagate to dependent Evidence Pages or learning assets where applicable, and reopen a verdict when genuinely new evidence changes the record.
Redacted Forge may earn money through eligible platform revenue, ads or Premium share, education-player licensing, contextual affiliate or shopping links, owned-audience offers, podcast surfaces, licensing, or future sponsorships when appropriate.
Commercial eligibility does not raise historical confidence. Paid relationships must be disclosed when applicable, and a commercial partner does not get to convert weak evidence into a stronger verdict.
Classroom packs, Courses, worksheets, source-literacy exercises, or educator use may be created from the same verified research. Redacted Forge does not claim accreditation, institutional endorsement, expert review, standards approval, or professional certification unless that specific status is real and documented.
For published investigations, the intended public chain is simple: claim → Receipt → source/locator → what it proves → limitation/counterevidence → verdict → correction history. Evidence Pages and “Did Redacted Check This?” claim pages are designed to make that chain reusable by viewers, educators, journalists, researchers, libraries, search engines, and citation systems without forcing them to trust the narrator alone.