Reports that Matter

About this edition

Assessing Russian Activities and Intentions in Recent US Elections

How the text was made, and where it falls short

← Back to the report

How the text on Reports that Matter was made, and where it still falls short of the printed page. Nothing has been rewritten. Where we know the text differs from the printed report, this page says so.

Last reviewed 24 October 2026.

The edition

  • The document: the Intelligence Community Assessment ICA 2017-01D, "Assessing Russian Activities and Intentions in Recent US Elections", dated 6 January 2017, as the Office of the Director of National Intelligence (ODNI) published it at dni.gov/files/documents/ICA_2017_01.pdf. The copy kept in the report's repository is the Internet Archive's capture of that file on the day of release (25 PDF pages, 1,589,106 bytes, SHA-256 5876e67e…2257); every complete capture of the address since then, to September 2024, is the same file. The PDF is the canonical citation target: page numbers on this site are its printed page numbers.
  • What it contains: ODNI's three-page note "Background to 'Assessing Russian Activities and Intentions in Recent US Elections': The Analytic Process and Cyber Incident Attribution", which opens the file, and the assessment itself: Scope and Sourcing, Key Judgments, Contents, the CIA/FBI/NSA assessment "Russia's Influence Campaign Targeting the 2016 US Presidential Election", Annex A (a 2012 Open Source Center report on RT) and Annex B (Estimative Language).
  • Licence: a work of the US government, in the public domain in the United States.
  • Size: about 6,560 words in 140 paragraphs, 1 note, 18 printed pages marked (i–iv and 1–14).

How the text was made

The PDF was produced by software (Aspose.Pdf) from a word-processed original, so its words are reliable, but it loses the document's structure: its bullets are drawn in a symbol font whose character the text layer does not identify as a bullet, its headings are marked only by type size and weight, the two-column pages of the assessment interleave in reading order, and Annex B is printed as a picture with no text at all.

  • Structure from Wikisource. The volunteer transcription of this same PDF on Wikisource (every page with text marked "Validated", that is, proofread by two people; the copy it was made from is byte-identical to ODNI's) gives the headings, paragraphs, bullets, bold and italic, the note and the contents. A copy is kept in the repository under reference/raw/, pinned by SHA-256. Its words are not taken on trust: every word is matched to the same word in the PDF, in order. 95% of the transcription's words find their place (98% of the PDF's words on the pages it covers); the rest are Annex B, below. Each paragraph is given the printed page its first word sits on, and a page that begins inside a paragraph is marked after it.
  • One correction to the transcription. Annex B on Wikisource reads "the intelligence Community's judgments"; the printed page has a capital I, and this text follows the page.
  • Annex B (page 13). The page is a picture: its words are not in the PDF's text, so they cannot be checked against it the way the rest of the report is. They were transcribed from the page image and compared word for word with Wikisource's validated transcription, and the transcription is recorded in the repository's corrections.yaml. The likelihood chart's two rows of terms ("Almost no chance" to "Almost certainly", "Remote" to "Nearly certain") are given as a small table, left to right as printed; the chart's colour scale and its axis (Percent, 0 to 100) are not reproduced.
  • Bullets. The assessment sets many of its judgments as bulleted paragraphs. Each bullet is served as a paragraph of its own, opening with "•", so that each has its own link and can be cited and quoted on its own.
  • The declassification banner printed at the head of every page ("This report is a declassified version of a highly classified assessment; its conclusions are identical to those in the highly classified assessment but this version does not include the full supporting information on key elements of the influence campaign.") is left out of the text as page furniture. The same statement appears in the report's own words in Scope and Sourcing (page i), where it is kept.
  • Page numbers. The assessment numbers its front pages i–iv and its body and annexes 1–14. The Background note prints its own page numbers, 1 and 2; they are left off here, so that a page number on this site always means the assessment's own page. The cover, the blank pages and the title page of the assessment's body carry no number.
  • The note. Annex A's title carries an asterisk note ("This annex was originally published on 11 December 2012 by the Open Source Center, now the Open Source Enterprise."). The site numbers notes, so it is shown as note 1.
  • Headings. Annex A and Annex B are headed as the contents names them ("Annex A: Russia -- Kremlin's TV Seeks To Influence Politics, Fuel Discontent in US", "Annex B: Estimative Language"); the page prints the annex letter and its title on two lines. The run-in bold headings of the assessment ("Cyber Espionage Against US Political Organizations.") are kept at the start of their paragraphs, as printed.

Known limitations

  • Photographs and charts are not shown. Annex A's five photographs (stills from RT and REN TV, and RT's editor in chief with Vladimir Putin) are not reproduced; their captions are kept, in italics, after the paragraph they stand beside. The full-page chart "TV News Broadcasters: Comparative Social Media Footprint" (page 11: YouTube views and subscribers, Twitter followers, Facebook likes and chatter for RT, Al Jazeera English, BBC World and CNN International) is not reproduced; the page is marked but has no text. Annex B's likelihood chart is given only as its terms (above).
  • The cover pages are not reproduced as text: the Background note's cover (with the seals of ODNI and the National Intelligence Council), the assessment's cover ("ICA 2017-01D | 6 January 2017") and the title page of its body (with the seals of the CIA, the FBI and the NSA).
  • Hidden text on the cover. The PDF's cover carries text that is not visible on the printed page: a second version of the banner ("This report is a downgraded version of a more sensitive assessment; its conclusions are identical to those in the more sensitive assessment…") and a classification line, "TOP SECRET//HCS-P/SI-G//ORCON/NOFORN/FISA". Neither is part of the visible document, and neither is in the text here.
  • Annex B's words are checked against the page image only, not against the PDF's text, which has none for that page (above).
  • Annex B shares a page of the site with the end of Annex A, because the site gives a short part its own page only above a minimum length; it keeps its own heading and its own page number, 13.

Reporting a problem

If the text here differs from the printed report, the PDF is the authority. Open an issue on the report's repository with the page number and the passage.