This is the third note in a series updating release information in NARA’s JFK Database. The first note is here and the second note is here
The first note explained basic issues with the JFK Database, looked at how NARA posted ARC documents online, and went over “release totals” and “record totals” at NARA from 2017 to 2026.
During this period, NARA posted ~~ 74,600 files online, representing 37,225 distinct records. Of these records, 37,213 were released in full (RIF), while 12 records are still available only with redactions in them. Most of these redactions are minor.
There are also a very small number of records in the ARC which were never posted online. The most notable of these records is the microfilm duplicate of Lee Oswald’s CIA 201 file. Note that the hardcopy 201 file has been released released in full.
The microfilm copy of the 201 file is divided into about 178 records in the JFK Database. None of these records were ever posted at NARA.
The second note in this series looked at how many records were “released in full” (RIF) for each of the 17 release sets at NARA. Using this information, we gained an overall picture of ARC releases at NARA under Trump 1.0, Biden, and Trump 2.0.
Of the records posted at NARA, Trump 1.0 released in full 60 percent, Biden released in full 33 percent, and Trump 2.0 released in full 7 percent.
This third note will look at the finer details of how redactions were released. There was much confusion over some of these releases, and a detailed look sheds light on the problems and pitfalls of complex declassification projects, one of the main topics of this blog.
Release dates and record counts
The key information from note two was a table listing how many records were released in full (RIF) on each of the 17 record releases at NARA. Rather than force readers to flip back and forth between this note and note two, I will copy the table again here.
The table includes the following basic release information: 1) release number, 2) release date, 3) total number of records released on that date, 3) records released in full (RIF) on that date, and the percentage of RIF releases versus the total number of records posted on that date.
| # | rel date | total rel | RIF rel | pct RIF |
|---|---|---|---|---|
| #1 | 2017-07-24 | 3793 | 3788 | 99% |
| #2 | 2017-10-26 | 2891 | 2888 | 99% |
| #3 | 2017-11-03 | 676 | 95 | 14% |
| #4 | 2017-11-09 | 13213 | 0 | 0% |
| #5 | 2017-11-17 | 10744 | 8342 | 78% |
| #6 | 2017-12-15 | 4256 | 1783 | 42% |
| #7 | 2018-04-26 | 18897 | 5519 | 30% |
| #8 | 2021-12-15 | 1491 | 1483 | 99% |
| #9 | 2022-12-15 | 13235 | 8834 | 67% |
| #10 | 2023-04-13 | 422 | 422 | 100% |
| #11 | 2023-04-27 | 355 | 355 | 100% |
| #12 | 2023-05-11 | 502 | 501 | 99% |
| #13 | 2023-06-13 | 290 | 288 | 99% |
| #14 | 2023-06-27 | 1106 | 228 | 21% |
| #15 | 2023-08-24 | 21 | 5 | 24% |
| #16 | 2025-03-18 | 2565 | 2549 | 99% |
| #17 | 2026-01-30 | 139 | 133 | 98% |
I will repeat here the caveat I gave last time: ARC docs were often released incrementally, meaning not all redacted material in a record was released at one time. Some records had as many as 5, even 6 different releases, with more of the record text becoming available each time.
The table above does not count these incremental redactions, which were numerous and sometimes complicated. Instead, it simply gives a count for all records RIF on that particular date. It is therefore a simplified record of the release process.
The rest of this note will go through the releases year by year. As the table shows, release years were 2017 (six releases), 2018 (one release), 2021 (one release), 2022 (one release), 2023 (six releases), 2025 (one release), and 2026 (one release).
2017 releases
The 2017 releases were the most extensive and complicated. This note will not look at record content. It will simply look at how and when records got released in full.
Based on RIF counts, the six release sets from 2017 can be divided into three groups: Releases 1 and 2, releases 3 and 4, and releases 5 and 6.
Releases 1 and 2 went as scheduled under the JFK Act. The July release was actually several months ahead of schedule, while the October release came promptly on 26 October, the date stipulated in the JFK Act. A total of 6676 records were released in full in these two sets. Virtually all records were RIF; fewer than 10 records were RIP.
Following the second release in October, President Trump signed an executive order postponing release of remaining redactions for another six months, while agencies continued to review their records.
Despite this order, however, NARA and the departments and agencies with redacted material, continued to post material online Some records posted in this period were RIF, but most were RIP.
The next two sets, releases 3 and 4, were posted at NARA in early November. These releases totaled almost 13,900 pdfs; only 95 were RIF. The other 13,000 plus pdfs each had at least one redaction, most considerably more.
It is worth noting that over 97% of these records were from the CIA. In fact, putting aside the CIA records RIF in sets 1 and 2, sets 3 and 4 constitute almost all of the other CIA records released by NARA from 2018 to 2026.1The NARA releases included 15953 distinct CIA records in total. Releases 1 and 2 included 2249 distinct CIA records.
Again, this doesn’t mean that no redactions were released in sets 3 and 4. It just means that there was at least one redaction box left in the pdfs posted.
In contrast to sets 3 and 4, Releases 5 and 6 were about 93 percent FBI records. Unlike the CIA records, a substantial number of these records were RIF: about 78 percent in release 5, and 40 percent in release 6.
As we saw in note one, before the 2017-2026 process began, redacted FBI docs actually exceeded redacted CIA docs. After releases 5 and 6, however, CIA docs had by far the majority of redactions in the ARC, a situation that held true all the way up until 2026.
These FBI releases in sets 5 and 6 are the main reason that 2017 had by far the largest number of records RIF.
2018 releases
On paper, set 7 from 26 April 2018 was the largest posting in the entire 17 set series. When one checks for RIF records, however, it was by no means at the top.
This was not more generally noted at the time. Why? Most of the CIA records from releases 3 and 4, and many FBI records from releases 5 and 6, were posted again in 2018, leaving researchers with little choice but going through over 18,000+ pdfs again if they wanted to know exactly how many records were RIF in 2018. As far as I know, no one did this.
This problem was not resolved until NARA published an updated version of the JFK database in 2021. The updated version identified which records were released in full in 2018.
To add to the confusion, over 1100 records were posted at NARA in 2018 which were not released at all in 2017. Several hundred of these records were from the Kennedy Library, and most of these JFKL records were omitted from the JFK database, confusing the picture even more.
The result of all this confusion was to thoroughly obscure the fact that of the 19,000 pdfs posted at NARA in 2018, only about 5500 were RIF. The others were either RIP, or, in some very confusing cases, duplicates of records that had been released in full in 2017. Small wonder that researchers were confused about how many redactions remained in the Collection at this point.
2021 release
The 2021 releases (set 8) were the first from the Biden administration. This abbreviated set (1491 pdfs) was an attempt to clear the decks of records that departments and agencies no longer wished to bother reviewing for release.
A review of these pdfs shows that many of them were duplicates of records that were already RIF in 2017-2018. Why weren’t these duplicate records also released in full in 2017-2018? This was obviously due to a hasty, careless review done under the gun of Trump’s six month postponement.
When set 8 first came out, I assumed all these records were RIF; I later found that a small number had redactions continued in them, despite the express intent of Biden’s memo.
NARA also posted 70 plus records not posted in the previous seven sets of records released under the Trump administration. All of this added to the continued confusion.
2022 releases
The 2022 releases (set 9) included 13235 pdfs, and were accompanied by detailed lists of records with redactions, explanations of the basis for the redactions under the JFK Act, and “transparency plans” which spelled out future reviews of the redactions and conditions for their release. These releases also included another 473 previously unposted records.
Of the 13235 records posted by NARA, 8830 were RIF, making 2022 the top RIF posting of all 17 doc sets published on NARA’s website. This was NOT generally acknowledged in most discussions of the 2022 releases.
Why? Discussions of set 9 often focused on individual records that individual researchers were interested in which were not RIF in 2022. This focus on individual records, rather than on overall redacting in the ARC, has been a constant feature of internet discussions since 2017.
This was of course due in part to difficulty of knowing what was RIF and what was not, but after the JFK Database was updated in 2021, this should have been clear.
After much confusion, most researchers came to realize that a large number of records had indeed become RIF, but few if any noted the actual number of RIF records.
This left about 4500 records only available online with redactions, though in most cases these were minimal. This drop from over 13000 records redacted to 4000+ meant that over 99 percent of the JFK ARC was RIF.
2023 releases
Set 9 in 2022 was followed by six smaller releases in 2023 (sets 10-15). All of the pdfs posted in 2023 were RIF, except for set 14, posted on 27 June 2023. Set 14 included 1129 records, only 228 of them RIF.
These six sets from 2023 were generally ignored in online discussions of the ARC status. Even after set 15 was posted, estimates of how many RIP records were still in the ARC varied from 4500 (ignoring the 2023 releases completely), to 3500, which seems to be a random guess rather than a real tabulation.
2025-2026 releases
Sets 16 and 17 in 2025-2026 are the latest releases so far, and come very close to disclosing all remaining redacted material in the ARC. The only document in this set which was not released in full is the CIA’s history of it’s Mexico City Station, which appeared in set 17 with some material still redacted.
This is one of 12 records posted at NARA which have not yet been released in full. There are also the Oswald microfilm records, which never posted at all.
There are probably a handful of other records which were posted at NARA put had very small redactions that were never completely released; I have seen at least one of these which I am still looking for.
There are also possibly a small number of records, omitted from the JFK Database, which still have small redactions. I have found at least one such record recently posted online by NARA in their catalog. NARA is committed to scanning all documents in the JFKARC and posting them on-line. This will be a long process which I will continue looking at and posting about.
Finally, there are a substantial number of records which are exempted from full release by the JFK Act. These include records incorporating grand jury material, court sealed material, and records donated to NARA where donors limited release of some or all portions of the documents. I will have another post later on these records.
My two cents
There are still several topics I want to cover in this series, but this note is an important one, which I have been working on for a long time. I doubt that most people will care about the things I have written about here, but I feel that basic picture I have drawn is essential for understanding how this final stage of the JFK Act has been handled.
The JFK Act was a remarkable commitment to transparency in government, and despite how long it has taken, it has been carried through in a remarkably thorough fashion, far more thorough than many had originally expected.
The government’s implementation of the act has been severely criticized for many years, criticism that continues to this day. In writing this blog, I have often found that the implementation of the Act has been much more thorough than critics have claimed, and many serious errors and inaccurate descriptions of the ARC as posted at NARA.
More remains to be said, but I after years of looking very closely at it, I have found the JFK Assassination Records Collection to be an authoritative work which will remain the primary source for anyone looking at almost any aspect of the JFK assassination.