What the Documents Themselves Say

What the documents state

The National Intelligence Council assessment Foreign Threats to 2020 US Federal Elections (document 6, dated 19 August 2020, NICA 2020-06885D) states in its Key Takeaways that Russia, China, and Iran “have the capability to conduct such activities, although it would be difficult for them to… manipulate voting processes at scale and without detection.” It is, per its own scope note, a “downgraded version of a paper originally published on 19 August 2020,” and it “does not make assessments about the impact of these efforts on the US.”

The companion vulnerabilities assessment (document 8) assesses that “cyber operations targeting the electronic tabulation of results could delay results reporting from affected jurisdictions, potentially creating public uncertainty but probably not affecting the integrity of certified results,” noting that certified tabulation copies are “processed offline” and “stored independently” of public-facing result displays. It further assesses that adversaries “could… make false claims about their ability to manipulate US election infrastructure” as part of an effort to undermine confidence.

Internal caveats the documents note themselves

The assessments repeatedly frame their conclusions in the language of capability and probability (“we assess,” “probably,” “could”), and document 6 expressly declines to assess impact on the US. The remaining six documents in this package (for example the CIA notes at 1 and 2, and the emails at 4 and 5) are released with redactions and/or as scans; their released copies expose limited or no extractable body text, so they are cataloged by their released title and markings rather than summarized here.

What the documents state

The readable material in this package is primarily the drafting and coordination record. A declassified email chain (document 15, 5–7 October 2020) records the decision to publish a “NIC alternative analysis on china/election,” in which the drafters state their “lead, in short non-compartmented form, is that we assess that Beijing has taken some low-level, exploratory steps to denigrate the President and shape voter perceptions ahead of the election,” while noting they are “clearly labeling this as alternative analysis… the perspective of NIO Cyber and myself only.” Related NSA coordination emails (13, 14) discuss downgrading and cross-posting election-relevant material for the assessment. Titled but scanned products in this package assert (per their released titles) PRC analysis and collection of U.S. voter-registration and military data across multiple states.

Internal caveats the documents note themselves

The alternative-analysis email is explicit that its view is a minority/alternative position the drafters expected the broader community to dispute, and that the memo itself was to carry “a large text box laying out the mainline view.” This is a documented analytic disagreement, not a consensus finding. Many of the substantive products (9, 10, 21, 22, 23, 24, 25, 30) are released as image scans bearing only a “SENSITIVE GOVERNMENT AGENCY” cover marking with no extractable body text, and the multi-part summary (26–28) likewise has no text layer; their specific assertions are not machine-readable in the released copies and are not restated here beyond their titles.

What the documents are

This package is an FBI case file concerning an investigation into alleged fraudulent voter-registration activity in Michigan. By their released titles, it comprises witness- and worker-interview memos (for example 46, 47, 48, 50, 51, 52), fraudulent-registration records (41), a four-year case timeline (40), the original referral (43), declination memos (54, 55), a voter-fraud spreadsheet (56), and GBI Strategies materials (53).

Limitation stated by the released copies

Every file in this package is a scanned image whose released copy exposes only the “APPROVED FOR PUBLIC RELEASE” stamp as extractable text; there is no machine-readable body text. Accordingly, this catalog does not characterize the contents, findings, or outcome of the Michigan investigation beyond each document's released title and type. The presence of declination memos (54, 55) is noted as a released fact, not interpreted here.

What the documents state (and attribute)

The DHS summary “Preventing Alien Voting” (document 57) states that DHS “has initiated multiple investigations related to allegations of non-citizen voting and registration,” and reports figures it attributes to its own review — e.g., that review of “the first set of public voter files from states that have not utilized the SAVE system” found a stated number of noncitizen registrations across four named states, and that, “as of June 22, 2026,” 25 states had processed voter records through the SAVE system. These are the document's own figures, attributed to DHS review; they are presented here as the document's assertions.

The report “Recognizing and Addressing Threats to Statewide Voter Registration Databases” (document 58, July 2026) states in its own executive summary that it is an “unclassified overview of the threats to statewide voter registration databases” that “relies on reporting from the intelligence community, law enforcement, and state election officials.” It asserts that “recently declassified records revealed that China breached multiple state voter registration systems prior to the 2020 election” and frames the core risk as what can be done with stolen data (for example, obtaining absentee ballots), rather than direct vote alteration.

Internal caveats the documents note themselves

Document 58 is explicit that it is an unclassified, derivative overview built on other reporting rather than a primary intelligence finding, and it frames its purpose as informing state and local officials so they can implement “recommended mitigating controls.” Document 57's figures are self-attributed to DHS review of public voter files and SAVE-system processing; the document does not present them as adjudicated in court or independently audited.