I Read Every Declassified File From July 16th. Here Is What They Actually Say — And Where I Have To Correct My Own Side.
On the evening of July 16, 2026, the President of the United States addressed the nation from the White House and announced the immediate declassification of intelligence covering five areas of American election security. NBC declined to carry it live. ABC declined to carry it live. NBC News NOW ran it on a streaming channel and promised a special report afterward, which is the television equivalent of telling your wife you will get to the garage this weekend.
Part 1 of this series covered the Tranche One and Two. We pick it up here in Part 2 at Tranche Three.
— TRANCHE THREE: THE COVER-UP. WHERE THE REAL STORY IS. —
This is the section the networks did not want to carry. And of everything released, this is the tranche where I will go to the mat.
Set the table first. National Intelligence Council Assessment 2020-06885D, August 19, 2020. The coordinated, Tier A, whole-of-community assessment. It says foreign states will continue to use covert and overt influence measures to sway U.S. voters’ preferences, shift U.S. policies, increase discord, and undermine confidence in our democratic processes.
On Russia: a range of measures primarily to denigrate former Vice President Biden, with some Kremlin-linked actors seeking to boost President Trump’s candidacy on social media.
I am putting that in my own article. In full. Because I am not going to be the guy who only reads the paragraphs he likes. That is the exact behavior I am about to spend two thousand words condemning.
And here is a detail from that same assessment that I want my own side to sit with, because it is uncomfortable and true: Russian actors circulated narratives about voter fraud resulting from mail-in balloting, and called some U.S. primary voting processes “rigged” by the Democratic Party.
So Moscow was pushing mail-in fraud narratives in 2020. Does that make mail-in fraud concerns wrong? No — an adversary amplifying a true thing does not make the thing false, and this is a basic logic error people make constantly. But it does mean that anyone who cares about this issue has an OBLIGATION to be precise, because sloppiness on our side is a force multiplier for theirs. When you overstate, you are doing volunteer work for the Lakhta Internet Research troll farm. Free of charge. Enthusiastically.
On Iran: an influence campaign to undermine the President and divide the country, which the Supreme Leader probably authorized.
On Turkey — and nobody is talking about this one and I cannot understand why — Ankara’s efforts included COVERT FUNDING OF U.S. MUNICIPAL ELECTION CAMPAIGNS. A NATO ally covertly funding American city elections, in a coordinated IC assessment, and the entire American press corps looked at it like a dog looking at a card trick.
On China, the coordinated judgment: China prefers President Trump be defeated. And then: Beijing did not intend to try to affect the election. Beijing probably calculates that a concerted effort to influence the presidential election would risk backfiring and offer uncertain benefit.
Now the CIA reporting the President quoted, and he quoted it accurately. In mid-2018, Chinese Communist Party policy was to leverage all domestic and foreign elements opposed to the U.S. President in an effort to reduce his votes and make him resign or prevent his reelection. China analyzed which U.S. states and sectors supported him in order to target those sectors with tariffs and induce them to lobby against him. China sought to identify U.S. journalists who had reported negatively on the President and pay them to write more negative articles. By mid-2019, China assessed that directing resources toward political swing states was worth doing, because those states would be crucial in 2020.
The CIA’s own summary of the intent: the PRC’s ultimate goal with regard to the U.S. elections was for the U.S. President not to be re-elected.
And a President’s Daily Brief item from June 25, 2020: a senior Chinese official told a White House official that Beijing had derogatory information on him, in an effort to compel him to take a restrained approach, and threatened to release it via social media. The IC assessed the threat as CREDIBLE. The note adds that lower-level Chinese officials had been recommending the collection and use of “black materials” against perceived anti-China U.S. officials since at least January 2019, and that this was the most detailed and authoritative such recommendation the IC had seen.
Beijing tried to blackmail a sitting White House official. That is in a PDB. That, by itself, on any other night in any other decade, is a week of wall-to-wall coverage.
NOW THE FIGHT.
On October 16, 2020 — three weeks before the election — the National Intelligence Officer for Cyber and the Director of Election Threat Analysis published NICM 2020-09381. Its own first page says it offers an alternative perspective to the IC’s assessment, was provided to the IC for comment, and is NOT an IC-coordinated memorandum. Their claim: Beijing has taken at least some low-level, exploratory steps to undermine the President’s reelection chances by denigrating him and shaping voter perceptions.
Their own confidence level, in their own document: LOW TO MEDIUM. Their own stated reason: much of the available reporting would be consistent with either the mainline or the alternative view.
That is two officials being honest about the limits of what they had. I respect it. Write it down: they did not overstate. They flagged their own weakness in the first paragraph. That is how a professional does it.
The mainline IC view, as stated inside that same memo: Beijing has not deployed influence efforts intended to change the outcome of the U.S. presidential election.
Then the interagency knife fight. The State Department’s Bureau of Intelligence and Research objected to the graphic language, writing that the evidence for China NOT interfering is stronger than the drafters were acknowledging, that they should make a judgment call rather than leave the reader to read between the lines, and — the line every analyst should have tattooed somewhere — that absence of evidence should not be used to justify a conclusion that it could be happening but we just do not see it. INR added: China’s influence efforts to shift public opinion on particular policy issues are not the same as election interference.
Then FBI Counterintelligence, DAD Nikki Floris, December 30, 2020, formally commenting on the minority view: to state this as bluntly as the NIO does, without evidence to support this claim, is extremely misleading. And: citing the absence of evidence is an absolute red line for us.
She also objected that minimizing source descriptions meant clandestine reporting of dubious credibility, poorly identified sourcing chains, and minimal track records was being given the same weight as signals intelligence on identified senior Chinese diplomatic and intelligence figures.
That is a real argument. It is a GOOD argument. I am not going to pretend otherwise just because I do not like who made it.
AND NOW THE PART WHERE IT STOPS BEING AN ARGUMENT.
November 20, 2020. An NSA Strategic Intelligence Analyst writes to the drafting group: we have deliberately massaged our one pending PDB to avoid any direct links to the election. And: the massive report we have been waiting on has been chopped up into 13 reports that likely will not get out until late next week.
Thirteen. One report, cut into thirteen pieces, timed to dribble out after everyone stopped caring.
The National Intelligence Officer for Cyber’s reaction, same thread: this is a really good example, but far from the only one, of what I have been raising since the summer — that the IC is deliberately avoiding mentioning a connection to elections for NON-SUBSTANTIVE REASONS.
And the Director of Election Threat Analysis, in a line I would frame if I were a worse man: Their PDB is not going to tie to the election? The mind boggles.
Then a December 23, 2021 email. The NIO for Cyber flags that a later report attributes to the Chinese military a unit which, in 2020, the IC had said it did not know the identity of — and now calls it an election-influence unit. His point: it is the same personnel doing the same kind of activity. We should consider how this looks, and on what analytical, non-partisan basis the IC makes calls one way or the other.
Same unit. Same behavior. Different label. Different administration.
Here is my read, and I am going to be precise about what is mine and what is theirs.
There are two available interpretations of the massaging, and the documents preserve both. Interpretation one: the mainline agencies were correctly refusing to overstate a link the evidence did not support, exactly as INR and FBI CI argued. Interpretation two: officials were deliberately stripping elections language for reasons that had nothing to do with the evidence.
The documents DO NOT adjudicate which is right. I want that stated flatly, because I am not going to hand a fact-checker a free win.
But here is what does not require adjudication, and here is what I am willing to defend against any comer:
While that fight was raging inside the building — while the NIO for Cyber was writing that elections language was being stripped for non-substantive reasons, while an NSA analyst was writing the word “massaged,” while thirteen chopped-up reports were sitting in a queue — the American people were being told, from podiums, on camera, without qualification, that this was the most secure election in the history of the country.
Both things cannot be true.
Either the fight was real, in which case “most secure in history” was a statement nobody was in a position to make. Or the fight was not real, in which case the NIO for Cyber and the Director of Election Threat Analysis were lying in internal correspondence for no reason.
Pick one. I will wait.
That is the scandal. Not “they stole it.” THEY DECIDED WHAT YOU WERE ALLOWED TO WONDER ABOUT. And then they called you a lunatic for wondering about it anyway.
A CLAIM I CANNOT VERIFY AND WILL NOT REPEAT AS FACT.
The President stated that an official inside the FBI wrote that she was running a “shadow government” to keep intelligence about China’s election meddling from becoming known.
I have gone through the released material available to me and I have not located that line in a form I can cite with a page number and a full context. A fragment resembling it does appear in the internal Skype and message reconstructions, in a thread that reads — at least in the portion I can recover — like an exhausted executive complaining about report-routing paperwork at the end of a very long day.
I am not saying the President is wrong. I am saying I cannot show you the page, and I do not build on ground I cannot show you the page for. Same with the burn bags. He said significant numbers of burn bags were found in formation, material that was supposed to be incinerated and never was, and that he believes it was gross incompetence rather than intent. That may be exactly right. It is not in the documents I have. So I am flagging it and moving on, and if the page shows up tomorrow I will write about it tomorrow.
That is what a lab notebook is for. You write down what you measured. Not what you hoped.
Verdict on Tranche Three: SUSTAINED, WITH A DIFFERENT INDICTMENT THAN THE ONE HE CHARGED. The documents do not prove the IC hid a stolen election. The documents prove the IC had a savage internal fight about whether to tell you anything at all — and then, while that fight was unresolved, the government told you everything was perfect.
— TRANCHE FOUR: THE REPORT THEY PULLED BACK —
This is the one everybody is going to screenshot. So let us be adults about it.
In late August 2020, FBI Albany produced an Intelligence Information Report — serial 4 212 7305 20 — disseminated around September 25, 2020 to all field offices and the wider intelligence community. The subject line, recovered from the recall register: Chinese Government Production and Export of Fraudulent US Drivers Licenses to Chinese Sympathizers in the United States, In Order to Create Tens of Thousands of Fraudulent Mail-in Votes for a Presidential Candidate.
That is the claim. That is where it came from. One report.
Now the sourcing, printed on the face of the document: this is an information report, NOT FINALLY EVALUATED INTELLIGENCE. It has not been fully evaluated, integrated with other information, or interpreted. Do not take action based on a presumption of accuracy.
And the source description: a collaborative source with indirect access, NONE OF WHOSE REPORTING HAS BEEN CORROBORATED, for less than one year. The source got it from an identified sub-source, who claimed to have gotten it from unidentified PRC government officials.
Unidentified Chinese officials, to a sub-source, to an uncorroborated source, to the FBI. That is third-hand at best.
I am a former paramedic. If I brought a doctor a history that came from a guy who heard it from a guy who heard it from somebody who would not give a name, that doctor would not intubate on it. He would not order labs on it. He would look at me like I had lost my mind.
It was recalled within hours. The reconstruction shows a DAD messaging a colleague — another one went out re: election security — and getting back “oh crap,” followed by “how’s the sourcing on the new IIR?” By 5:09 p.m.: totally concur to pull this back until we get back in front of this guy. By 5:17: contacted by Cyber DAD and CD DAD, they are directly requesting it be recalled until the source is reinterviewed.
The recall register is the detail that will not leave me alone. Roughly 1,130 FBI recalls from January 1, 2019 to the present. Albany’s total substantive recalls for the entire year 2020: ONE. This one.
Internally, agents wrote: personally, I am highly skeptical of this reporting. And: portions regarding COVID that veer into conspiracy. The Foreign Influence Task Force China group did not want it reissued, and their stated reason was substance, not politics — one message says explicitly it is nothing to do with partisanship, it is legitimately the information, and we would be trying to drill down regardless.
So. Anybody telling you that document proves China manufactured tens of thousands of ballots for Joe Biden is handing the other side a loaded weapon, pointed at us. It was raw. It was uncorroborated. It was third-hand. It was recalled. The people who handled it did not believe it. Stop it. Just stop.
BUT.
Buried inside all of that is the single most damning artifact in the entire release, and almost nobody is going to read far enough to find it.
An FBI agent objected. In writing. On the record. To his own chain of command.
His argument, reconstructed from the handout: he was not satisfied with the reasoning behind requiring a new source report and a headquarters review. What most concerned him was the repeated statement that the reporting would CONTRADICT DIRECTOR WRAY’S TESTIMONY TO CONGRESS. He wrote that this implied to him that one of the reasons they were not putting it out was POLITICAL — and that this cuts directly against the Bureau’s mission to remain apolitical and simply state what it knows. He wrote that it is not the field’s job to assess whether intelligence aligns with the Director. The field provides intelligence so that people way above their pay grade can make the call.
He acknowledged the other side of it, too — that they are obligated not to put out reckless information or something they know to be false. He is not a crank. He was wrestling with it honestly, in real time, and he wrote his objection down anyway, knowing it would be filed.
Whoever you are, sir: I do not know your name. But you did the right thing when the right thing was expensive, and I have been in rooms where that mattered.
Because look at how many times the phrase appears in that file. It contradicts Director Wray’s testimony. Again on the next page. And the next. And the next. And the next. Five separate times, across pages 19, 20, 21, 24, and 29 of the same handout.
Now — the FITF China group’s stated reason was substance. I believe that they believed it. And it is entirely possible for BOTH things to be true: that the intelligence was garbage AND that some people in the building were relieved it was garbage because publishing it would have made the Director look like he had misled Congress.
That is not a conspiracy theory. That is human nature, and it is why we have process. When the same justification for suppression appears five times in one file, and that justification is “our boss said something different in front of Congress,” the process broke. Full stop. The reporting may still have been junk. The reason cited was contaminated.
The record also shows that a redrafted version was ultimately DELETED due to “internal coordination,” and that a note went out to all field offices requiring that ALL election-related raw reporting must henceforth route through FBI Headquarters — a requirement that, per the file, did not exist the week before.
Read that in sequence. A report about election interference goes out. It gets recalled. Then a new rule appears requiring that every future report about election interference be cleared by headquarters first.
I am not going to tell you what that is. I am going to let you do what my students do when they see a data point move right after somebody adjusts the instrument.
AND ONE MORE THING, BECAUSE HONESTY IS THE WHOLE PRODUCT HERE.
The 19,800 counterfeit U.S. driver’s licenses seized out of Chinese shipments at Chicago — that thread is real and it appears in the file as an actual investigative lead, with agents identifying China-based sender names and shipping addresses. And I wrote about it back in my article “A Senator Vows to DIE Stopping Voter ID. I Read the FBI Affidavit. Let’s Talk.” I said the barcodes worked. They tested them in Michigan. They scanned clean.
But the file contains a detail I did not have then, and I am putting it in my own article against my own interest: CBP’s own assessment was that the MAJORITY of those documents are used by teenagers between 18 and 20 to get into bars.
That is what the government thought they were for. It is in the file. It probably explains most of them.
Does that dispose of it? No. A fake ID with a working barcode does not care why it was manufactured. Its function does not change based on the intent of the guy who ordered it. The Chicago seizure is one port, and the thing that has always bothered me about the “it’s for beer” explanation is the volume — but I do not have the evidence to push past bothered, and bothered is not evidence.
Verdict on Tranche Four: THE HEADLINE CLAIM IS NOT SUPPORTED. THE HANDLING OF IT IS AN OUTRAGE. The report is junk. The five-times-repeated reason for burying it is the story.
— TRANCHE FIVE: MUSKEGON, MICHIGAN. THIS ONE MAKES ME ANGRY. —
October 29, 2020. FBI Detroit opens a case — 56-DE-3249380 — after the Muskegon City Clerk’s Office reports that beginning around October 5, it received six packages by U.S. Mail containing voter registration applications. The clerk determined some were fraudulent based on non-existent addresses, invalid telephone numbers, and signatures that did not match Michigan Secretary of State voting or DMV records. And many of them appeared to be in the SAME HANDWRITING.
Michigan State Police executed search warrants on October 29, 2020, including at the Best Western Grand Haven hotel. The Michigan AG, MSP, and Muskegon PD were all in it. The FBI notified both U.S. Attorneys.
The volume, per the later DOJ summary: approximately 8,000 to 10,000 applications mailed from the same company. And when workers came in person to drop off more, those had the same problems.
During the MSP search, agents found boxes of “blue cards” — reloadable prepaid debit cards. Subsequent interviews indicated the company used those cards to pay workers who brought in voter applications.
Then the database checks. This is the part where I stop being funny.
Of 107 applications examined: NINETY-ONE returned no results whatsoever. Not a bad address. Not a typo. Ninety-one human beings who did not exist. Of the sixteen who were confirmed to be real people, only FOUR had signatures on file that matched the signatures on the application.
Four out of one hundred and seven.
Of twenty randomly selected applications with apparently different handwriting: seven completely fabricated. Four with a real name but a mismatched date of birth, driver’s license, or Social Security number. One with a real name and the wrong middle name. One for someone who never lived in Muskegon, at a fabricated address. Seven that were correct or nearly correct.
Individual entries, and I am reading these straight off the page:
Number 2. Unable to locate anyone by that name. Address does not exist. Phone number does not return to anyone.
Number 14. Name located. DECEASED — BUT WAS ALIVE AT THE TIME OF REGISTRATION.
Number 19. Unable to locate anyone by that name and date of birth. THE ADDRESS IS A PARK.
The address is a park.
I have written this sentence before. I wrote it in “Is There Voter Fraud in Fulton County? Here Is What Bski Knows,” published February 24, 2026, where I documented independent journalists using a forty-five-dollar voter roll and Google Maps to find active registered voters at a park, a MARTA bus station, a jury duty parking lot, an abandoned nightclub, a vacant lot demolished in 2007, and a sixty-bed homeless shelter with 205 registered voters.
I got called a conspiracy theorist for that article.
An FBI agent wrote “the address is a park” in an official memo in 2023, and the United States Department of Justice filed it.
The canvasser interviews. One worker said she provided information because it was EASY MONEY. She said she was never directed to make things up and never did it herself, but that she heard about people submitting falsified forms so they could turn in as many applications as possible. Another said he would just sit in his car the whole shift and put down the name of a friend or a relative in order to have something to turn in with his clipboard.
Sitting in a car. Inventing Americans. Getting paid on a prepaid debit card.
THE TIMELINE IS WHERE YOUR BLOOD SHOULD BOIL.
October 27, 2020: MSP briefs the FBI Detroit Election Team.
March 11, 2021: the DOJ Public Integrity Section authorizes a full field investigation into the PREPAID CARDS. And then this line, which I would like carved somewhere: INVESTIGATION INTO VOTER FRAUD NOT AUTHORIZED.
They greenlit the debit cards. They said no to the voter fraud.
August 12, 2021: PIN did not approve further action in the Western District.
November 4, 2021: formal declination.
February 2, 2023 — two years and four months after the packages arrived — PIN finally authorizes a full field investigation, but only on a narrowed scope. Only applications where the purported applicant does not exist or appears fabricated, AND where signatures are the same or substantially similar to signatures on other applications. The concurrence explicitly does NOT extend to applications featuring the names of actual persons with a unique signature.
October 3, 2023: PIN invokes the Department’s election non-interference policy, Justice Manual section 9-85.300, and its election-year sensitivities policy, section 9-85.500 — reasoning that interviewing canvassers close in time to the November 7, 2023 election MAY CREATE A PUBLIC PERCEPTION OF THE DEPARTMENT AND FBI ACTING WITH A PURPOSE OF AFFECTING THE OUTCOME of that upcoming election.
I want to sit in that one for a second.
The Justice Department declined to interview witnesses in a voter registration fraud case because doing so, before an election, might look like the Justice Department was trying to affect an election.
The same Justice Department that ran Operation Arctic Frost against nearly 100 Republican organizations, including Charlie Kirk’s Turning Point USA, as documented in the Senate Judiciary Committee record. Apparently the appearance concern is a seasonal item. It goes on the shelf in the off-months.
December 12 through 15, 2023: mass interviews finally conducted. THREE YEARS AND TWO MONTHS after the fraudulent applications hit the clerk’s desk.
March 11, 2024: the AUSA advises he does not expect this case will lead to prosecution. The agent requests a written declination.
May 15, 2024: the FBI Laboratory’s Questioned Documents handwriting report is completed and forwarded to the AUSA with a request for a charging decision.
They had the handwriting analysis. From the FBI Lab. On applications where the clerk’s own initial complaint was that they were all in the same handwriting.
Nothing.
And the agent — again, an agent, again, in writing — objected. He wrote that he was not comfortable closing the case without raising these issues, and that he was not thrilled with the wording, because it stated that PIN concurred with our decision not to pursue it, as opposed to PIN DIRECTING US NOT TO PURSUE IT.
Read that again. He is documenting, for the file, that the paperwork was being written to make a top-down kill look like a bottom-up decision. He knew exactly what he was looking at and he made sure it was written down.
Two men in this entire record did their jobs. Both of them are anonymous. Both of them were overruled. Both of them wrote it down anyway.
NOW THE HONEST PART, AND IT IS A BIG ONE.
The DOJ manual that the FBI agent himself quoted — Federal Prosecution of Election Offenses, December 2017, pages 42 and 43 — says this about paying people per registration: while this situation USUALLY DOES NOT RESULT IN FRAUDULENT VOTES ACTUALLY BEING CAST, it does cause registration offices to become overloaded, and risks overloading voter rolls with “deadwood” names, which undermines public confidence.
Registration fraud is not the same thing as vote fraud. Ninety-one fictional people on applications is not ninety-one fictional ballots. The FBI never established that fraudulent votes resulted. The agent himself acknowledged that proving the conduct was WILLFUL — the actual legal standard — would still be difficult.
There is one file whose TITLE indicates a voter confirmed she did not register but received a ballot. The substance of that memo is redacted. The title is all there is. So it is a lead. It is not a finding. I will not pretend otherwise.
The President said the FBI agents believed crimes were committed and the DOJ slow-walked it and killed it.
That is accurate. Both halves. The agents did believe crimes were committed — they cited the statutes, 52 U.S.C. 20511(2) and 10307(c), and noted that 10307(c) eliminates the need to prove a given pattern of corrupt conduct actually impacted a federal race, because the harm is presumed. And the Department did slow-walk it, for four years, across two administrations, invoking a policy about the appearance of interference in order to avoid interviewing witnesses.
But he left out the part where the same manual says this kind of fraud usually does not produce fake votes. And that omission is exactly the kind of thing that lets a fact-checker throw the whole thing in the trash.
Verdict on Tranche Five: SUSTAINED AS TO CONDUCT AND SUSTAINED AS TO THE BURIAL. NOT ESTABLISHED AS TO VOTES CAST. Ninety-one fictional people. Debit cards. A four-year slow roll. A handwriting report nobody acted on. And the paperwork massaged to make the kill look voluntary.
Mike Borowski is a medically retired Army combat medic with 23 years of service, including a combat deployment to Iraq, and a high school Anatomy and Physics teacher at a high-need Career Technical district in Northeast Ohio — where he also wrote and published the textbooks for both courses. He runs “Bski’s Classroom,” a platform dedicated to cutting through political noise with data, history, and the kind of blunt honesty that comes from someone who has seen both war and the American classroom up close.
X (Twitter): x.com/BskiMike22802
Substack: substack.com/@bskimike22802
YouTube: www.youtube.com/@BskiMike22802
AFNN: https://afnn.us/author/mikeborowski/
If you enjoyed this article, then please REPOST or SHARE with others; encourage them to follow AFNN. If you’d like to become a citizen contributor for AFNN, contact us at managingeditor@afnn.us Help keep us ad-free by donating here.
Substack: American Free News Network Substack
Truth Social: @AFNN_USA
Facebook: https://m.facebook.com/afnnusa
Telegram: https://t.me/joinchat/2_-GAzcXmIRjODNh
Twitter: https://twitter.com/AfnnUsa
GETTR: https://gettr.com/user/AFNN_USA
CloutHub: @AFNN_USA