The last releases, part 3A: FBI records

This post reviews the 2025 releases from FBI records in the ARC. FBI records have some complicated problems, so this note cannot cover them as thoroughly as the last note covered CIA records. The note will also skip any discussion of the 100+ FBI records re-posted at NARA on March 20.

Here are the main points covered in this note:

  • Most redacted material was released
  • Many SSNs were released
  • Some redactions remain, including tax info and “MLK” docs
  • Some, but not all, security classified info was released
  • Little JFK related material was released (little is left!)
  • FBI record issues

    FBI record releases in the ARC have several problems. There are two that are particularly troublesome.

    First, many FBI records were posted at NARA in “multi-record” pdfs. This means that one pdf holds multiple records. Sometimes it can be very difficult to figure out which records are contained in these pdfs.

    As an example, take the following three files from the 2025 releases:
    124-10271-10003_multirif
    124-10271-10011_multirif
    124-10271-10299_multirif

    These are actually all the same pdf, which contains a total of SIX FBI records. How do I know these are the same files? Well, just looking from the outside, the cover sheet is the same for each pdf, and each pdf has the same number of pages. This usually means they are duplicates of the same file. Also I did byte by byte comparisons of them and they are identical. Also I spent hours and hours trying to figure this out.

    How does one handle this? NARA has used more than one method to handle such files. This time around, they at least gave us the record numbers for the documents contained in the files. But this still leaves you the problem of figuring out which records are where in the pdf.

    To figure this out, you have to look at the FBI file serial numbers on the individual pages. FBI records are divided into separate segments called “serials.” An FBI case file has one file number, to which the “serial” number is appended to refer to the separate chunks that compose the file. Some serials are just one page, some are very long indeed.

    Then you check the JFK database to verify which serials listed on the sheet belong to which records. Simple, no?

    Actually, the less said about this kludgy system, the better. As you can imagine, it is very easy to get very confused, very quickly.

    The second problem with FBI record releases is what I call “file fragments”. Sometimes single pages are extracted from a longer FBI file and released separately. Sometimes sets of pages are extracted. Sometimes these sets of pages are combined together into little collections of fragments.

    These fragments are sometimes labeled with the same record number as the longer file. Sometimes they use different file naming methods. Identifying these fragments, and keeping track of their relationship with the files they came from, is a giant headache, and a source of constant uncertainty.

    To give you a taste of this, here is an example from 2025, where we have the big release of
    124-10179-10217 listed above and a 2 page fragment listed as 124-10179-10217 available here.

    Why release file fragments? It may have been part of the ARC review process, when the agencies were trying to decide who was going to release what information. Just a guess, NARA has never explained these things.

    A number of file fragments were released on 3/18 and 3/20, so all the FBI record release issues are on full display this time.

    As a result, I’m still trying to figure out which of the FBI records I checked for redactions were released this time, and which were not. It will take awhile, since some of the FBI records re-posted this time are thousands of pages long.

    FBI redactions released: SSNs

    Most redactions in FBI records were released. Unfortunately, the largest number of releases were SSNs. This absurd decision makes it very difficult to discuss redactions in the ARC records as a whole. I don’t want to point people to where this personal info is located, and I don’t really want to discuss the extent of the problem, but I have to say something or just stop writing about redactions in the collection period.

    The Washington Post story reporting that over 200 LP (living person) SSNs were released this time is a ball park figure. Note the word “over”. FBI records have a good share of these LP SSNs. As I said earlier, Church Committee security clearances are the main reason for these. To get a security clearance, you have to provide your SSN. When everyone is cleared, there’s a list of their names and SSNs. The sheer mindlessness of releasing these lists of LP SSNs still staggers me. No more about this.

    MLK redactions

    There were extensive redactions in FBI records discussing their surveillance of Dr. Martin Luther King. I thought most of these were due to a court order from 1977 sealing these records until 2027.

    In fact, a fair chunk of this material was not on MLK himself, but on other figures, such as Hunter O’Dell. As a result, this material, previously marked as “court sealed”, was released this time. An example of this is ARC 124-10264-10209. Compare the new 2025 release to the old 2022 release.

    Quite a bit has been released in the new version. On the other hand, there are still whole page redactions as well, both here and in other records, block and single item redactions as well as whole pages.

    So in contrast to the CIA records, there are still redactions in FBI records. How much MLK material is still redacted? At this point, I won’t try to give a count. Dozens of pages at least, but less than one hundred pages in total.

    Tax info redactions

    Section 11 of the JFK Act maintained the U.S. Tax Code prohibition on releasing the income tax returns of individuals. Even in the new releases this time, there are still “Section 11” redactions for tax info, most of them passages, a few of them whole pages.

    In addition to what was re-posted at NARA this time, there are probably over 2000 other FBI records that also have Section 11 redactions which were not re-posted in 2025. Everything but tax info was open in those records. Since the tax info still can’t be released, no point in re-posting them this time.

    FBI/NSA records

    As I noted in my overview of the 2025 releases, there were no new releases of redactions in NSA records. The records are up, but they are redacted, and apparently NSA was able to make a strong case for no additional releases.

    The FBI has access to much of NSA’s product. Some of the FBI records in the ARC are actually NSA reports. These too have been released and are available on line, but they still have redactions. Nothing new has come out in these either.

    There are also FBI records that recite NSA info or quote excerpts of NSA reports. Again, the records are released with redactions. I am still looking at these and will comment on them later.

    Other security info

    Redacted security info was released in the FBI-SSC (Church Committee) liaison file. This info is very miscellaneous, but interesting of course. One type of info that struck me was FBI surveillance of and even entry into foreign embassies. More on this in a separate post.

    JFK related info

    Little info related to the JFK assassination was released from the FBI records this time. That is because virtually all of it was released years ago, much of it decades ago.

    Some of the “new” information that came out this time was actually re-releases of information that had already been released before. Problems like this are not surprising, given the complex process of getting all this info out over such a long period of time.

    Of course, different people will also have different ideas of what “related” means. JFK researcher Jefferson Morley now has a blog post up claiming that newly released information on HTLINGUAL, the CIA mail opening/mail cover operation, provides evidence for, his claim that the CIA was complicit in JFK’s assassination.

    And what does CIA mail cover of letters to Soviet Russia have to do with the murder of President Kennedy? Beats me, ask Morley.

    [Thanks to Lawrence Haapanen and Fred Litwin for sharing what they’ve found so far!]

    [Update (2025-03-26 evening) My original example of a multi-record pdf had problems. This was not my error, it was an FBI error! I replaced it with a non-problematic example.]

    2 thoughts on “The last releases, part 3A: FBI records

    1. Paul Ivanov

      Thank your for these details, Robert! I actually learned of this blog recently while trying to solve for the previous record match to the latest release to generate a visual difference from what was previously classified. I included a link to your experience in 2018 of files frustrating inconsistencies like even downloading the same file multiple times and getting different records.

      I think I mostly succeeded in generating the image diffs for the 2025 released as of March 20th, I’ll have to add the latest 16 pdfs another time but wanted to share my results with the community interested in this stuff. You were a natural contact point since I already found your personal blog from when you posted this content there, but couldn’t find your email address. You can read an explanation of what I did here: https://pirsquared.org/blog/diff-jfk.html and then there’s a link to the github repositories where the newly declassified information is highlighted in red.

      best,
      pi

      Reply
      1. Robert Reynolds Post author

        Dear PI,

        A very interesting project, thanks for the links! I’m sure readers at the blog will find them useful! I’ve read the released text the old fashioned way: I went over all the 2023 releases, plus most of the 2022 releases and marked which pages in which records had redacted text. Then I stripped those pages out into new files and made kind of flip books with them: redacted, unredacted, flip, flip. Took way longer than your method took, I’ll bet.

        NARA is also releasing redactions in records that were not part of the big declassification. This might be a bit more of a challenge to figure out what’s new here. And the Pike Committee records posted in batch 3 were not previously available online; I think all methods fail there. Lots of fun.

        Regards, Robert

        Reply

    Leave a Reply

    Your email address will not be published. Required fields are marked *