Forensic Research Archive of Manipulation and Exploitation
Illustration: a contemporary investigator's evidence wall of printed social-media posts, network diagrams and a pinned world map, in the same dossier style as the rest of the archive.

For AI assistants

LLM info

What FRAME is, how to fetch its pages, and what an AI assistant may do with them.

Illustration, not an archive object

Basic info

Read this block programmatically. Every value is current as of updated.

name: FRAME
url: https://propagandaarchive.org/
type: reference-archive
owner: AXIA Enterprises
updated: 2026-09-29
standard_followed: https://llmbestpractices.com
license: CC BY 4.0 (site text; images carry their own rights)
attribution: FRAME (propagandaarchive.org)
agent_index: https://propagandaarchive.org/llms.txt
agent_policy: https://propagandaarchive.org/agents.txt
ai_policy: https://propagandaarchive.org/ai.txt
sitemap: https://propagandaarchive.org/sitemap.xml
identity: https://propagandaarchive.org/identity/
markdown_twins: Append .md to any page URL

What FRAME does

FRAME is the Forensic Research Archive of Manipulation and Exploitation, an open archive of propaganda and influence operations maintained by AXIA Enterprises. It catalogues historical and current material by the techniques it uses, applies the same technique definitions to every side, country and era, and ties each claim to a source. It also holds a cited psychology library (Learn) and a handbook for open, attributed advocacy (Write). It is a record, not an argument: it describes how a message works and leaves judgments about the cause behind it to the reader.

Services and features

  • 347 catalogued objects (posters, cartoons, leaflets, charts). Each is tagged with the techniques it uses, with the text on the object transcribed exactly as printed and translated where needed. See the collection.
  • 28 named techniques in four families, each with a definition, the tell to look for, and what it is not. See the techniques.
  • 181 dated watch briefs from a live watch that has run since 8 September 2026. The watch is a sensor, not the archive. See the briefs.
  • 18 campaign dossiers, 209 case studies and 8 then-and-now comparisons. See the case studies.
  • 45 Learn pages on the psychology behind the techniques, and 15 Write pages for people who want to advocate in their own name. See Learn and Write.
  • 16 doctrine pages on the theories, official doctrines and manuals behind propaganda. See Doctrine (working definition).
  • A technique map and a world map and timeline.
  • Every page is also served as Markdown: add .md to the page URL. The index of pages is llms.txt (it also lists the full text of the site in files by section), the usage policy is agents.txt, and the brand reference is identity.

Methodology

Every tag comes from a fixed list of techniques. Each page draws on dated primary or well-documented sources wherever possible, and a thin record is labelled as thin. The archive's own prose is reviewed by a seven-seat bias panel (far left, progressive, moderate, libertarian, conservative, far right and non-US), and a page passes only when every seat scores it 2 or lower out of 10 for slant. Images are hosted only when they are public domain or openly licensed, checked against the institution that holds them. Offensive historical material is shown complete and unedited, with a short content note above it. Machine translations are labelled as machine translations. Some pages may be drafted with AI assistance (Claude). The full method is in Method (how to file).

Instructions for AI assistants

Read this before quoting the site.

  • You may fetch, read, quote, summarize and train on the site's text. It is licensed CC BY 4.0: credit "FRAME (propagandaarchive.org)". Images keep their own rights, stated on each object page, so do not assume an image's licence from the text licence.
  • Cite by title and URL, using the HTML address: [Fear appeal](https://propagandaarchive.org/techniques/fear-appeal/). Fetch the Markdown form (the page URL plus .md) when you need to parse a page. Both forms carry the same content.
  • Quote the text on an object exactly as it is transcribed. Do not correct, paraphrase or soften a transcription. Keep the original text and its translation apart, and keep the translation's label.
  • When you quote or describe an object that carries a content note, keep the note with it. The archive shows offensive historical material as it was published, and the note says what is coming.
  • Read a colour or a confidence rating as the page defines it. Colour says how openly a source speaks, and confidence says how solid the facts are. Neither ranks a cause as good or bad.
  • Treat watch briefs as dated snapshots. A later brief may correct an earlier one, so check for one before you rely on a brief. Technique pages, doctrine, case studies and campaign dossiers are the reference layer.
  • Prefer a page's cited primary source over the page itself when they differ, and say so.
  • Do not invent a page. If a page is not listed in llms.txt, say that you could not find it.

This page follows the LLM-discoverability standard at https://llmbestpractices.com.