What the Release Was

Listen to this chapter, read by the author →

People talk about the release as if it were an event, one afternoon in March 2025 when the files came out. It was not. It was six separate drops over ten months, and the shape of them tells you more about how this was done than anything in the documents. Thirty-one thousand pages. Then thirty-seven thousand more. Then fourteen thousand. Then fifty-three pages. Then seven hundred. And then, ten months later, eleven thousand more.^{1, 2, 3}

Now look at what time of day each one went out, because the National Archives publishes it. Seven in the evening. Half past ten that same night. Half past nine at night. Half past three. Seven in the evening again. Those are not the hours of an archive. Nobody schedules a considered publication of historical records for half past ten at night. Those are the hours of people working to a deadline somebody else set.^{1, 2, 3}

The pair I keep coming back to is the eighteenth of March. Eleven hundred files at seven in the evening, and another thousand at half past ten. Three and a half hours apart, on the same day, from the same collection. That is not two decisions. That is one batch that was not finished when the first half was pushed out the door, and somebody went back to the building and finished it.^{1, 2, 3}

Then the tail, which nobody reported because it is not dramatic. Fifty-three pages on the twenty-sixth of March. Seven hundred and four pages, in two hundred and seven separate files, on the third of April. Two hundred and seven files averaging three pages each. That is somebody going back through the collection finding strays, one at a time, for weeks after the cameras left. It is the least glamorous entry in the whole record and it is the one that reads like actual archival work.^{1, 2, 3}

The final release, on the thirtieth of January 2026, does not match any of them. Eleven thousand pages, but only a hundred and forty files, which is about seventy-nine pages each. Every earlier batch averaged between three and thirty-five. So the last thing to come out was not a pile of cables. It was a small number of very long documents, and long documents are exactly what a page-by-page review leaves until the end. It also went out at ten in the morning, on a business day, like something that had been finished rather than something that was due.^{1, 2, 3}

As for how it is arranged, the Archives states it plainly: the files are sorted by Record Identification Form number, which is in the filename. That number was not invented for this. It was assigned in the 1990s, when the review board processed the collection, and it encodes which agency held the paper and where in that agency’s own filing it sat. Which means the release is organized by custody. Not by date, not by subject, not by person. By who was holding it.^{1, 2, 3}

Think about what that does to you as a reader. There is no folder for Oswald. There is no folder for the motorcade, or Mexico City, or the autopsy. Everything about a single subject is scattered across agencies according to which office happened to file it sixty years ago. You cannot browse this collection by any question a human being would actually ask. You can only browse it by bureaucracy, which is the same reason nobody found the things in this chapter until somebody turned the whole pile into searchable text.^{1, 2, 3}

§There is a wrinkle in that worth knowing about. Sixty of the files do not follow the naming rule. Some carry a second internal agency number in brackets after the record number. Others are labelled multirif, meaning one document holds several record forms at once. So a file can contain a record you are hunting for while being named after a different one, and any search built on the standard name skips it silently. They are only findable by counting what you have against what the page lists.^{1, 2, 3}

And I want you to notice what that is, because we have met it before. Earlier in this book there are documents the Bureau kept out of its own index by hand and documents the Agency kept out with a printed checkbox, and the effect in both cases was a page that exists and cannot be found. Here is a release from 2025 doing it again by accident. Nobody hid those sixty files. They are simply unfindable by the identifier they are filed under, which for a researcher is the same thing.^{1, 2, 3}

Here is the part that matters most and gets reported worst. As of the eighteenth of March 2025 nothing in that collection is withheld for classification. Not one page. But the Act itself has two other sections, and the Archives cites them. Section ten covers anything under court seal or protected by grand jury secrecy. Section eleven covers tax return information under the Internal Revenue Code, and anything deeded to the government by a private citizen with conditions attached. Those were never classification questions, so declassifying everything did not touch them.^{1, 2, 3}

Which leaves the one this book has spent its length on. Court seal, tax law and a donor’s conditions are all written down, and you can read the section number that holds each of them shut. The fourth is not in the statute anywhere. It is the material that belongs to somebody else’s government, and there is no section to cite because Congress never wrote one. Everything else that is still closed is closed by American law. That one is closed by an arrangement nobody has ever had to publish.^{1, 2, 3}

So my reading of the whole thing, for what it is worth. It was real, it was enormous, and I do not think it was a performance, because you do not spend the following ten months quietly posting fifty-three pages here and seven hundred there if the point was the headline. But it was executed at a speed that shows in every part of it. Files pushed out at half past ten at night, sixty of them misnamed, and personal information about living people published because there was no time to check. It was a genuine opening, done in a hurry, arranged by whose filing cabinet it came from.^{1, 2, 3}

† disputed   ‡ theory   § reconstructed — everything unmarked is established. Superscript numbers key to the sources below. Corrections

Sources for this chapter

3 sources. Every claim above is keyed to one of them. The complete list for the whole book runs to 1,276.

1. National Archives, JFK Assassination Records 2025 Documents Release page
2. President John F. Kennedy Assassination Records Collection Act of 1992, sections 10 and 11
3. the reconstruction