The 2026 election declassification

What it is

A technical summary of CISA's 2019–2024 election security work: source-code and binary analysis of election software conducted with Idaho National Laboratory under the Critical Product Evaluation program, penetration testing of state, local, tribal and territorial networks, and incident response. Findings are that election software contains the ordinary spectrum of vulnerabilities; that certification regimes block timely patching, with some regimes requiring no patches for months before an election; and that SLTT networks lack basic segmentation, with assessors gaining full network control within hours or days in multiple cases. Cites Halderman's ImageCast X barcode analysis from Curling v. Raffensperger. Mitigation recommendations: harmonize patch and certification rules, use human-readable paper ballots, and conduct post-election manual audits before certification.

Why it matters

Written for this release and unclassified, but it cuts against the speech rather than for it. Its recommendations — paper ballots, manual audits — are the mainstream election-security consensus, not a case for the claims Trump made. Most usefully, it declines to endorse the Puerto Rico work: it states that ODNI commissioned a forensic examination of Dominion devices from Mojave Research, and that while CISA reviewed the report, the agency had no access to the devices and could not perform its own examination. That is the agency holding the pen refusing to vouch for the administration's centerpiece hardware finding.

What it is

National Intelligence Council memorandum on infrastructure vulnerabilities ahead of the 2020 election. Assesses that cyber operations against tabulation could delay results reporting and create public uncertainty but probably would not affect the integrity of certified results, because tabulated results are stored independently of the copies displayed on results websites. Notes risk-limiting audits, then required in 38 states. Cites the Pennsylvania Blue Ribbon Commission finding that transmission of preliminary results to public websites is vulnerable to man-in-the-middle attacks that would not alter the separately stored certified copies. Recommends physical and cyber hygiene, third-party vendor verification, public messaging, and private messaging to adversaries.

Why it matters

The most quotable document in the release, for reasons the White House cannot have intended. A full section is headed 'False Manipulation Narratives Could Undermine Public Confidence.' It warns that adversaries could make wholly fabricated claims — the example given is announcing they had compromised all US voting machines — which would be difficult, time-consuming, or impossible for the government to disprove; that much of the voting public knows little about election administration, which could let false narratives gain traction; and that a widely publicized compromise would undercut confidence even if nothing was manipulated. It recommends timely public messaging to make US officials the trusted source. Written in January 2020, it describes the 2026 event it was released to support.

Uploaded to Drive with a truncated 8.3 filename (NICM_V~1.PDF). Byte size matches the original; content intact.

What it is

The mainline National Intelligence Council assessment on foreign election threats, in a downgraded version. Judges that Russia, China, Iran and many non-state actors have the capability to compromise election infrastructure but would find it difficult to manipulate voting at scale without detection. On Russia: assesses it is using a range of measures primarily to denigrate Biden, directing or encouraging proxies to spread claims about him, with some Kremlin-linked actors seeking to boost Trump's candidacy on social media; names Derkach and Kilimnik among proxies advancing a Burisma corruption narrative, with the stated aim of defeating Biden and ensuring Trump's victory. On China: assesses Beijing prefers Trump be defeated and has stepped up critical public rhetoric, but did not intend to try to affect the election. Scope note explicitly states the paper makes no assessment about the impact of these efforts.

Why it matters

Released under Trump's own name, and it contains the Russia-preferred-Trump finding in plain language. Also supplies the baseline that every other China document in the release is measured against, and the reason the dissent exists at all. The scope note matters for anyone trying to read impact into it: the assessment declines to make one.

What it is

The minority view itself, written by the National Intelligence Officer for Cyber and the Director of Election Threat Analysis. Header reads 'ALTERNATIVE ANALYSIS.' Scope note states it offers an alternative perspective to the IC's assessment that China has not attempted to influence the outcome, and that it 'was provided to the IC for comment before publication but is not an IC-coordinated memorandum.' The authors assess Beijing took at least low-level, exploratory steps to undermine Trump's reelection chances through overt messaging, nascent online covert influence capability, diplomatic measures and economic leverage; state these probably did not include Beijing's most aggressive options; and record low-to-medium confidence in their own assessments. They state they agree with the large majority of the IC's judgments about China's calculus. The memo reproduces the mainline view alongside: that Beijing has not deployed influence efforts intended to change the outcome, prefers stability in the relationship, and wants to preserve the ability to repair relations afterward. It concedes that much of the available reporting would be consistent with either view.

Why it matters

This is the source Trump quoted. The White House released it with every disqualifying marking intact — the alternative-analysis banner, the not-IC-coordinated disclaimer, the low-to-medium confidence, the authors' own concession that the reporting is ambiguous and that they agree with most of the mainline. The gap between what the speech said this document shows and what the document says about itself is the story, and it is legible from the first page.

What it is

Email from Nikki Floris, then Deputy Assistant Director of the FBI's Counterintelligence Division, sent to the Director of Election Threat Analysis and copied to Tonya Ugoretz, commenting on the updated ICA draft. The FBI is fine with a minority-view textbox but insists its analysis meet the same analytic standards as the rest of the ICA — 'an absolute red line for the FBI' — and opposes putting the minority view in the Key Judgments, where without context and source descriptions it would be given parity with the ICA's assessments. Point by point: on the NIO's claim that China's leaders 'intended' to affect candidates, the IC has repeatedly stated on a body of intelligence that there is no evidence of such intention, and stating it that bluntly is 'extremely misleading.' On sourcing, minimizing source descriptions gives clandestine reporting of dubious credibility, poorly identified sourcing chains and minimal track records the same weight as SIGINT on identified senior Chinese diplomatic and intelligence figures. On the claim China 'may' have encouraged protests, there is no evidence in the reporting and the citations do not support it — 'a misrepresentation of the underlying reports.' On the dissent's collection-gaps argument, citing the absence of evidence is 'an absolute red line for us.' The forwarding note records that the FBI had previously said Director Wray signed off on the ICA text.

Why it matters

The single most damaging document in the release, and the administration published it. It is a named senior FBI official's contemporaneous, itemized finding that the analysis Trump built a primetime address on failed the IC's own tradecraft standards. Floris is on the record and reachable. Note also that this sits in Section 01 rather than with the China material in Section 02, which is either filing drift or a deliberate burial — either way, worth asking about.

What it is

Email from the National Intelligence Officer for Cyber — the minority view's author — a year after the ICA. He argues that a new IC report cites some of the same intelligence and logic used in 2020 to conclude that particular Chinese units were not engaged in election influence, and now uses it to argue that China IS influencing a (redacted) election. He notes the report attributes activity to 'the Chinese military' when in 2020 the IC said it did not know who it was, and characterizes as an election-influence unit the same unit he and a colleague said was influencing the 2020 US election. He asks on what logical, non-partisan basis the IC makes calls one way or the other, and says it should be highlighted for oversight.

Why it matters

The dissenter's strongest argument, in his own words: an inconsistency claim about how the IC applies attribution standards across cases. It is a real methodological complaint and shouldn't be waved off — but it is an argument about IC consistency, not evidence that China altered a US election, and it does not answer Floris's sourcing objections.

What it is

CIA World Intelligence Review item on Chinese cyber activity around the campaign. Reports that China is probing the presidential campaign to tailor collection on US-China policy positions, and that Chinese cyber actors have done so in every US presidential campaign since at least 2008. Since 2018, actors known in the private sector as APT31 have targeted personal email accounts of senior US leadership including officials in the Executive Office of the President, Congress and the federal judiciary. As of 20 May, APT31 sent spear-phishing emails with tracking links to Gmail accounts of campaign staffers; on 4 June, Google announced APT31 was targeting the campaign, and Google and the FBI briefed campaign officials, with Google stating publicly that the attempts were unsuccessful. Records that during the past year Chinese cyber actors collected US election-related information from voter databases, a polling data company, political and nonprofit organizations, fundraisers and campaign advisory organizations.

Why it matters

Documents real Chinese cyber espionage against the Biden campaign — and then states the conclusion the speech needed to avoid: the IC assesses that China does not currently intend to covertly interfere to try to sway the outcome of the election, although the activity could enable such operations if Beijing decided to. Note the target was Biden's campaign, not Trump's, and Google says the phishing failed. Another Warrington-track release that contradicts the Trump-track framing.

What it is

A 2026 CIA retrospective summarizing selected 2004–2020 reporting on Venezuelan interest in manipulating electronic voting systems. Finds that Venezuelan officials developed sustained interest and likely some capability in manipulating electronic voting systems including Smartmatic technology; that the 2006 IC assessment that Smartmatic's Venezuela ties posed a national security threat rested on solid intelligence about Venezuelan intent to influence US politics and evidence of manipulation of Venezuela's own systems; and that this led the US to force Smartmatic to divest its US operations in 2007. Cites April 2004 reporting that Chávez stated an objective of preventing the reelection of a sitting US president.

Why it matters

Reads as a setup for a Dominion/Smartmatic revival and then refuses to deliver. Its own summary states the intelligence did not definitively confirm that large-scale electronic fraud was successfully executed in specific Venezuelan elections, that CIA's baseline assessments maintained other factors better explained electoral outcomes, and that reporting on the more advanced techniques came from limited sourcing. Everything concrete concerns Venezuela's domestic elections and a US divestment order completed nineteen years ago. Carries a distinct third declassification stamp — Ratcliffe, 1 July — with Warrington's release approval added 10 July.

What it is

The White House Government Transparency Task Force's cover statement. Records that Trump created the task force within the Executive Office of the President in May 2026 to advise him, through the Counsel to the President, on documents that should be declassified or released 'to further his administration's priorities.' Announces the first disclosure, states that declassified intelligence reveals voter registration rolls from at least 18 states have been compromised by the PRC, and that additional records reveal more than 200 million voter records were compromised 'without state-specific affiliations.' Names sixteen jurisdictions: Alaska, Arkansas, Colorado, Connecticut, District of Columbia, Florida, Georgia, Iowa, Kansas, Maryland, Michigan, New York, North Carolina, Ohio, Oklahoma, Rhode Island.

Why it matters

The document that turns intelligence reporting into a political claim, and the one to check the rest of the section against. Three problems on a single page. It says 'at least 18 states' and names sixteen jurisdictions, one of which (DC) is not a state. It concedes the 200 million figure has no state-specific affiliation, which means the headline number cannot be tied to any named system. And the task force's stated purpose is not transparency in the abstract but furthering the administration's priorities — the charter is in the text. Ohio's presence on this list is not supported by the document it derives from (see 'US Voter Registration for 6 States'), which records that the Ohio download failed.

What it is

One-page summary asserting that ahead of the 2020 election China had extensive plans to use influence and cyber operations to sway opinion against the Trump administration, designed to exploit US societal fissures across several scenarios including US-China military conflict. Themes listed here: economic recession; encouraging violent demonstrations and looting to increase the appearance of social unrest; intensifying disapproval of the President and calling his ability to govern into question; undermining the legitimacy and public support of the President; ineffective USG COVID-19 response. States China had developed capabilities to project these themes into social media (TikTok, Facebook, Twitter, YouTube) and mainstream media through overt and hidden influencers, and that one option included gathering information on senior US officials.

Why it matters

Read the verbs. 'Had extensive plans,' 'had developed capabilities,' 'one of the options included' — this is a capabilities-and-options document, not a record of executed operations, and it contains no voter data at all despite being filed in the voter-data section. Parts 1, 2 and 3 share an identical opening paragraph and identical closing paragraph, differing only in the bulleted theme list, which suggests a single source product split into three releases rather than three findings.

What it is

Identical framing to Part 1. Themes listed here: racial tensions, including alleging that White people hate Black people, inciting demonstrations and marches as evidence of racial divides, increasing conflict between police and anti-racism activists; USG tensions with the US military; political party disagreements; Congress–Trump administration conflicts; women's rights; US South China Sea policies; alleged US war aspirations; international propaganda.

Why it matters

Same structural point as Part 1: planned themes and developed capabilities, not documented operations or effects. The 2020-era mainline assessments in this release address exactly this question and conclude Beijing did not deploy them to change the outcome.

What it is

Identical framing to Parts 1 and 2. Themes listed here: gun proliferation; immigration policies, including inciting migrant community dissatisfaction with the USG, instigating demonstrations against alleged USG human rights violations against immigrants, inflaming resentment between anti-immigrant and pro-immigrant communities, and instigating anti-immigration demonstrations.

Why it matters

Completes the set. Note that the three parts together describe a menu of influence themes spanning both directions of most US political arguments — which is what a capabilities catalog looks like, and is hard to reconcile with a claim of a targeted operation to defeat one candidate.

What it is

Reporting on PRC analysis of US voter registration information from 18 states, and plans to conduct US person matching and public opinion analysis. Describes voter registration records from the midterm elections containing personally identifiable information, to be used for mining PII on US citizens and as a resource to analyze, discover and identify US citizens; states the information would be used to conduct US general election analysis, and that the actor intended to continue obtaining state voter registration data.

Why it matters

The apparent source of the White House's '18 states' figure. What it describes is analysis of voter registration data for identity matching and opinion research — the ordinary use case for a commercial voter file. Redaction removes how the data was obtained, which is the only question that matters for the word 'compromised.' The companion six-state notice answers it: commercial websites.

OCR partial. Heavy redaction; acquisition method not recoverable from the released text.

What it is

Near-duplicate of the 18 States Memo, covering PRC analysis of US voter registration information from multiple states to conduct identity matching and public opinion analysis, dated 2019. Same substance: records from the midterm elections, containing PII, to be used for mining PII on US citizens, analyzing and identifying US citizens, and conducting analysis on the US general election, with intent to continue obtaining state voter registration data.

Why it matters

Substantially overlapping text with the 18 States Memo, which inflates the apparent volume of the release. Same limitation: the acquisition method is redacted. Note the filename typo ('Analsyis') is the government's, not an upload artifact.

OCR partial.

What it is

Reporting that a PRC actor shared 2020 US voter registration data for cities within seven US states in 2023, and separately inquired whether a counterpart had data on Connecticut and Massachusetts.

Why it matters

2023 activity involving 2020-vintage data — three years after the election the speech is about. Nothing here bears on 2020 vote casting or counting.

OCR partial. State list beyond Connecticut and Massachusetts is redacted.

What it is

Reporting on PRC collection of US consumer, military and voter registration databases and capabilities to conduct identity verification checks of US persons. Lists a US consumer database, US voter registration databases from state governments including Georgia and Iowa, and a registered US military personnel information database, described as building a comprehensive PII baseline on US persons.

Why it matters

The clearest statement of what this material actually is: bulk PII aggregation for counterintelligence identity resolution — the OPM/Equifax/Marriott pattern — with voter files as one commercially available input among consumer and military data. A real and serious espionage concern, and an entirely different thing from election interference. Georgia and Iowa here are two of the sixteen the White House names.

OCR partial. Record counts redacted.

What it is

Reporting from 2023 on PRC discussion of targeting the 2024 US elections. Includes a formal correction notice. Key points section is almost entirely redacted in the released version.

Why it matters

Filed in a section about 2020 voter data and concerning 2024. Too heavily redacted to characterize further — I can confirm the subject and the presence of a correction notice and nothing else. Do not let the filename do analytical work the released text cannot support.

OCR partial. Substantively unreadable as released; subject line and correction notice only.

What it is

Reporting that a PRC entity, seeking US presidential election-related intelligence, collected publicly available information from a US government website. The material collected is described as 2020 US presidential election infographics and products — maps and visualizations of each state's mail-in voting procedures, a mail-in voting policy change map, post-election and election-returns procedures — intended to provide election officials, interested parties and voters with information, and products setting out procedures to counter rumors and safeguard the 2020 election.

Why it matters

China downloaded CISA's public voter-education infographics. That is the finding. It is filed under 'Acquisition and Exploitation of American Voter Data' and stamped for release by the Counsel to the President. If you need one document to characterize the evidentiary standard of this section, it is this one.

Uploaded with a truncated 8.3 filename (PRCUSP~1.PDF). Byte size matches; content intact.

What it is

CIA note collecting portions of sensitive reporting disseminated between 2018 and 2020 on the PRC, explicitly stating it is not a comprehensive synopsis or summary of all reporting from that period. Includes a mid-2018 report that the PRC's ultimate goal with regard to US elections was for the US President not to be re-elected, and that China found the uncertainty of the administration unsettling.

Why it matters

Where Trump's 'as far back as mid-2018' line comes from. Preference is not interference, and the IC's mainline assessments in this same release say so directly: Beijing preferred Trump lose and chose not to act on it. Note the self-limiting scope note — this is a selection, not a survey, and the selector was the task force.

OCR partial.

What it is

PDB item titled 'Beijing Escalating Efforts To Shape US Policies on China.' Reports that Beijing told a White House official it held derogatory information on him in an effort to compel a restrained approach, with a threat to reveal it through social media, and assesses the threat is credible. Notes that lower-level Chinese officials have recommended collecting and using 'black materials' against perceived anti-China US officials since at least January 2019, and that this is the most detailed and authoritative such recommendation seen.

Why it matters

A genuine and serious counterintelligence item — attempted blackmail of a named-but-redacted White House official. It is about shaping China policy, not elections, and it is filed in the voter-data section. Heavy redaction; the substance above is most of what survives.

OCR partial. Two pages, majority redacted.

What it is

Continuation of the same matter three days later. A senior official separates two issues: first, the PDB that has apparently been worded away from elections — they have not seen the draft, but the characterization is familiar from a few weeks earlier, and they would be concerned as to why analysis is being 'deliberately massaged' away from elections; second, a broader NSA effort to consolidate reporting for the 45-day report. Another participant writes that they are not sure what to make of it, and asks why, if the item is off, the NSA officer says 'ok'. The underlying NSA email, dated 20 November, reads in relevant part: 'OK, thx. Regarding cross-posting into various [redacted] or downgrading to [redacted] for the 45 day piece, if you make sure there are [redacted] versions, we're good.'

Why it matters

The counterweight to the 20 November chain, and the reason to be careful with it. The NSA email actually in the record is a routine note about cross-posting and downgrading for the 45-day piece — it says nothing about avoiding elections. Senior recipients are visibly unsure the allegation holds up, and one asks why the NSA officer would say 'ok' if the item were being killed. A suppression story needs the PDB draft itself, which is not in this release.

OCR partial.

What it is

Coordination chain around the draft alternative analysis, nine days before the minority view published. A CI subject-matter expert in CIA's East Asia division writes that the draft 'speaks exactly to what has been on my mind but what was left out of the mainline IC assessment' and that it looks very good. The Director of Election Threat Analysis forwards it. A recipient responds that an NSA senior joining is 'a big deal' because he is 'one of the few seniors that remained analytically focused.'

Why it matters

Useful for the sociology rather than the substance: it shows the dissent being assembled and its authors recruiting supporters, with a working assumption that colleagues who disagreed were not analytically focused. Corroborates that the alternative analysis was circulated for comment — consistent with the memo's own scope note — while remaining un-coordinated.

OCR partial.

What it is

Coordination chain on a four-country election security graphic. The National Intelligence Officer for Cyber writes that he is concerned about politics seeping into the product, that there is no reason the IC cannot state its confidence level and that adversaries have a preference, and that while he does not disagree the consensus view is most likely, he strongly disagrees the IC can be so confident about it; he adds that he is troubled by CIA's comments. Another participant asks whether the IC is 'basically saying we should self-censor our assessment on preference because it could be misused,' notes that the actors do have a preference and that the IC assesses they think acting on it is probably not worth the risk, and says that if the IC cannot communicate its thinking clearly, 'maybe we should just quit and go home.'

Why it matters

The dissenters' complaint in its most sympathetic form — an argument for stating preference and confidence plainly rather than softening findings against misuse. Note what it concedes in passing: the participant arguing hardest against self-censorship states the assessment as adversaries having a preference but judging action not worth the risk. That is the mainline view, from the dissent's side of the table.

OCR partial. 24 pages, largely signature blocks and repeated classification headers.

Uploaded with a truncated 8.3 filename (EMAIL_~3.PDF). In the White House download this file sat at the section root rather than in the numbered subfolder.

What it is

The FBI Albany Intelligence Information Report itself, date-time group 25 September 2020, sent from Director FBI to all field offices and a very long external distribution including CIA, NSA, DIA, State, DHS, DOJ, Treasury and the combatant commands. Bates-stamped FBI-SJC-IIR-000001, indicating production to the Senate Judiciary Committee.

Why it matters

The primary document in the recalled-IIR controversy, released with its distribution list intact — which is itself the point, since the recall's significance turns on how widely it had gone before being pulled. The Bates prefix shows this reached Grassley's committee through congressional oversight, so parts of this story are already on the record and datable.

OCR partial. The released pages are dominated by the routing header; substantive body text is largely redacted.

What it is

Byte-identical duplicate of the Section 01 copy. See that entry for the substance.

Why it matters

Filing duplication. Counts toward the release's apparent volume without adding a document.

What it is

Byte-identical duplicate of the Section 01 copy. See that entry for the substance.

Why it matters

Filing duplication.

What it is

Byte-identical duplicate of the Section 01 copy. See that entry for the substance.

Why it matters

Filing duplication.

What it is

FBI Detroit opening electronic communication for the full field investigation, filed under STATE OF MICHIGAN; CDPO – ELECTION LAWS, marked Sensitive Investigative Matter.

Why it matters

Administrative but load-bearing for chronology: it dates the formal opening to 16 March 2021, five days after PIN's authorization. Establishes that the matter was opened, not buried.

OCR partial.

What it is

Email from FBI Detroit's Grand Rapids resident agency transmitting the spreadsheet 'Person Checks_2023.xlsx' detailing information and outcomes of person searches, with a second tab named 'Stats' providing percentage breakdowns.

Why it matters

A transmittal email. The spreadsheet itself — the actual data, with the percentage breakdowns — is not in the release. Given that the whole matter turns on match rates across 107 applications, the withheld attachment is more informative than the email announcing it. Worth a FOIA.

What it is

FBI Detroit EC recording searches conducted on 22 March 2022 in Accurint, Clear and other commercial databases against a named individual from the applications. Filed under STATE OF MICHIGAN; CDPO – ELECTION LAWS, Sensitive Investigative Matter.

Why it matters

Routine investigative record. The filename says 'Fake Voter'; the document is a records check on a name from a registration form. Three of these are in the release, which is worth noting when counting documents.

What it is

Companion EC to the above, same date, same database searches against a different named individual.

Why it matters

Routine investigative record. Near-identical to Searches EC (1) apart from the subject name.

What it is

Third EC in the series, recording searches conducted 5 May 2022 in Clear, Accurint and through the Michigan Secretary of State.

Why it matters

Routine investigative record. Notable only in that the check ran through the Secretary of State's own systems — the state cooperating with the federal investigation the release implies was being stonewalled.

What it is

FBI Guardian complaint import form, filed to the zero file for election law matters, recording a call to the FBI on 8 December 2020 reporting possible election fraud in Southfield, Michigan. The complainant describes an unknown subject wearing a Biden shirt coming to her door on 3 November 2020.

Why it matters

A single unverified public tip, filed to a zero file — the FBI's holding bin for matters not opened as investigations. The filename asserts a voter confirmed she did not register but received a ballot; the released text does not establish that, and the disposition is not in the release. This is the weakest document in the section and the one most likely to be quoted as though it were a finding.

OCR partial. Do not characterize beyond the complaint's existence and filing status without the underlying Guardian record.

What it is

FD-302 of a canvasser interviewed in December 2023. In fall 2020 she was approached outside an office next door — previously an eyeglass store — asked if she wanted a job, applied to register people to vote, and was hired the following day. Describes the office as run by a tall thin white male.

Why it matters

One of eight canvasser interviews conducted after PIN lifted the election-sensitivity hold. The substance is a description of getting hired at a storefront operation. Names and the operative allegations are redacted; the filename is doing considerably more work than the released text supports.

OCR partial. Heavily redacted.

What it is

FD-302 of an interview conducted at the subject's residence on 7 July 2025 by FBI special agents. The recoverable content is biographical: the subject works as a certified nursing assistant and as a medical transportation driver, with her shift patterns recorded.

Why it matters

The most inflammatory filename in the release and the least supported by its own text. Nothing about gift cards or suppressing Trump survives in the released portion — the substance is redacted. Note the date: July 2025, nearly two years after the canvasser interviews and squarely inside the period the task force was assembling this release. Worth asking who directed a fresh 302 in July 2025 on a matter DOJ declined in 2021, and whether the interview series that produced it is complete in this file.

OCR partial. Allegations in the filename are not verifiable from the released text. Do not repeat the filename as a finding.

What it is

FD-302 of a telephonic interview. In 2020 the subject was employed by an organization that assisted in registering people to vote; she was given a T-shirt and forms, could not remember the organization's name, was referred by a friend, and was told she would be paid.

Why it matters

Establishes the pay-per-registration structure that gives canvassers a motive to fabricate — the recognized failure mode of paid canvassing, and the reason most states regulate it. The 'others falsified' half of the filename is not recoverable from the released text.

OCR partial.

What it is

FD-302 of a canvasser who heard about the job through her uncle, wore a green and blue shirt, reported to an office that used to be an eye care center, canvassed for registrations outside a Walmart, and returned to the office at the end of each shift.

Why it matters

Same storefront operation as the other 302s. Nothing in the released text corresponds to the filename.

OCR partial. Filename claim not verifiable from released text.

What it is

FD-302 of a telephonic interview with a canvasser employed in 2020 by an organization that registered people to vote. Shifts were four hours; she reported to an office building where she was directed where to solicit registrations, and spent most shifts outside gas stations.

Why it matters

Nothing about other states is recoverable from the released page. On its face this is a description of shift logistics.

OCR partial. Filename claim not verifiable from released text.

What it is

FD-302 of a canvasser who initially could not recall working for a company involved in the 2020 election and remembered only when reminded of where he applied — near a Dollar General. He recalled wearing bright neon clothing and working only a couple of weeks in the fall.

Why it matters

A witness with weak recall of two weeks of gig work three years earlier. Nothing in the released text establishes forgery.

OCR partial. Filename claim not verifiable from released text.

What it is

FD-302 of a canvasser interviewed by two special agents on 12 December 2023. He worked for the organization in 2020 and was tasked to find people who were not registered to vote and to do whatever he could to get them registered.

Why it matters

The quota pressure, in a canvasser's own words. Bates-numbered rather than editorially titled, which is worth noting: the documents given descriptive filenames are the ones the task force wanted read a particular way.

OCR partial.

What it is

FD-302 of a canvasser who worked between September and October 2020, Monday to Friday 8:00am to 12:00pm, while caring for an elderly person. She was contacted about the job, went to an eyewear store near a Dollar General, provided her driver's licence and was given a W-2 and other paperwork.

Why it matters

Same storefront, same recruitment pattern. The W-2 detail matters: this was documented employment, which is why the FBI could reconstruct the canvasser list at all — and why a wholly covert scheme is a poor fit for the facts.

What it is

Department of Justice / FBI letterhead from Grand Rapids, Michigan, file 56D-DE-3407960, dated 25 September 2025. Addressee and body redacted.

Why it matters

Undatable in substance but useful in fact: the case file was still generating correspondence in September 2025, four years after PIN declined it and two months after the July 2025 302. Establishes the matter was reactivated in the run-up to this release.

OCR partial. Body redacted; letterhead, file number and date only.

What it is

Email chain from 28 June to 12 August 2021. The chain opens with PIN's authorization to conduct a preliminary investigation and a request for a synopsis of findings to forward to PIN for further guidance. It closes on 12 August with word that PIN informally advised no further action should be taken in this district at this time, with an official response to follow, and a note that the status of any authorization for the Eastern District investigation was unknown.

Why it matters

Documents the first declination, in the ordinary sequence: authorize a preliminary investigation, review the synopsis, decline. Nothing in the chain indicates a reason other than the merits.