Redacted CIA docs, 1990-1999

This is the third note in a chronological review of redacted CIA docs from the ARC. For an overview of the dates of redacted docs, see the first note here. For a review of CIA docs from the 1940s that still have redactions, see the second note here. This note will review redacted docs from 1990 to 1999. These include the most recent docs in the ARC.

By far the majority of these documents are related to the work of the Assassination Records Review Board (ARRB), the independent federal board that oversaw the assembly and release of the JFK ARC.

Some of these documents cover long standing controversies over declassification. There are sometimes lengthy redactions remaining in these documents.

Others reflect the ARRB’s search for JFK-related documents that were not included in the Warren Commission and HSCA document sets. Some of these, dealing with sensitive topics such as liaison relations and communications intelligence, are also still redacted.

Particularly notable are CIA cables covering ARRB requests and questions. These cables are almost all redacted, for reasons I will look at below.

Finally, some of the CIA liaison personnel with the ARRB were also under cover, and their names are also still redacted. Some of these names may remain redacted well past 2022.

Standard disclaimer for all posts on ARC redactions: what is written on ARC redactions today will not necessarily be true tomorrow. Tracking redactions in the ARC is like aiming at a fast moving target, as NARA reviews the collection, updates its database, and prepares for more releases in December 2022.

Dates in the JFK database

The JFK database has a doc date field for every record, but it does not always put something there. Some records are inherently undatable: file markers, folders, tab pages, and the thousands of miscellaneous scraps of paper in the ARC naturally are not datable. The JFK database usually leaves the date field for these blank.

Other records are lengthy, compound documents which have multiple dates in them, and are often not arranged chronologically. The dates that appear for these compound documents are thus somewhat random and copies of the same document may well show different dates. Sometimes a single document may also have different, conflicting dates on it as well.

Finally, more than a few of the documents are simply dated wrongly. A dozen or so documents dated to the 1990s were in fact from the 1960s. I have left these out of my discussion; a detailed accounting would make this note not just unreadable, but unwriteable as well. As with my other posts, if there is any interest in seeing the detailed list of the records I counted, please leave a comment on this page.

ARRB and JFK-M documents

Documents related to the ARRB, as well as other JFK Act related matters, are classified in the JFK Database as JFK-M (m for miscellaneous I guess). To find these, check the beginning of the JFK database comments field for “JFK-M:”1ARRB docs are not the only subject covered in the JFK-M category. Another page is coming on CIA record categories in the JFK Database. More details will go there.

When going through the JFK-M records, it is very useful to count by the number of records under a disk number. The disk number is the first eight digits in the unique record number assigned to each record in the ARC. ARRB related records start with the disk number 104-10301 and go up to 104-10340. Each disk covers a specific topic, listed in yet another field in the JFK database called “file number”. Check the current status field for “Redact” records, and one can get a short list of how many records of which type are still redacted. I will note here again that NONE of the CIA records are withheld in full.

Following is a count of redacted records related to ARRB and the JFK Act organized by disk number:

disknum total # of records redacted records percent redacted description
104-10306 28 1 ~4% DCI files
104-10324 3 1 ~33% Halpern interviews
104-10326 104 101 ~97% Cables on ARRB/JFK Act matters
104-10328 39 1 ~3% ARRB determinations
104-10330 139 34 ~24% CIA-ARRB correspondence
104-10331 383 185 ~48% CIA-JFK Act correspondence
104-10332 25 20 80% Declassification issues
104-10333 18 16 ~89% ARRB matters
104-10335 17 10 ~59% ARRB formal requests
104-10336 39 25 ~64% ARRB informal requests
104-10337 15 11 ~73% CIA matters
104-10338 22 4 ~18% 3rd agency referrals

Adding all these records up should give a total of 409.

As samples, I have included a link to a redacted document for the description in each row. Where possible I have tried to pick records that have more redactions, as an example of the upper bounds of redaction. In a few of these, the redaction may not be so obvious. For example, 104-10328-10038 has only two redactions on the the first page, both the names of CIA staff. Nothing else is redacted in this 93 page list.

I should emphasize that this list does not include all of the JFK-M records. Specifically, it does not include records prior to 1990 and it does not list disk numbers where all the records are open. As noted above, the dating is straight off the JFK database, and may be wrong in some cases. There are also a handful of records which the updated JFK database lists as redacted, but which seem to be open in full (see for example 104-10338-10011). I have included these in the count.

The ARRB cables

The most notable set of redactions is of course in 104-10326, where almost all the records are still redacted. These are all CIA cables to overseas facilities (i.e. “stations”). Releasing these cables in full officially confirms the presence of CIA facilities in specific locations abroad. This is a major issue for some host countries, which are probably all “witting” of CIA facilities and personnel in their countries, but for powerful political and security reasons are firmly opposed to public acknowledgement of this. Naturally, acknowledgement also poses security and policy issues for the U.S. government.

Withholding the locations clearly marked on the cables, however, also poses a significant problem for the JFK ARC. Cables make up at least a third of all CIA records in the ARC, probably more. Much of the content of these documents is simply incoherent without the locations. In addition, some of the locations are easily identified, simply by the cable content. Withholding the locations in such cases is of course problematic.

One way CIA and the ARRB attempted to balance these issues was to open chronological “windows”. Early on in the ARRB’s release process, CIA agreed to release cable locations important to the assassination story such as Mexico City for certain periods, such as 1963-1964. As ARRB pushed for fewer and fewer redactions, first locations, then dates, expanded.

The ARRB cables represent the farthest limit of this expansion, since they all date to the mid- or late 1990s. It is an open question whether these documents will be released in full by December 2022. If host countries pose strong objections, President Biden will have to decide whether to take more criticism in the U.S. or to risk damaging liaison relationships abroad.

Note that there are still documents with redacted locations from the 1960s, the core period of the collection. My interpretation is that the countries involved have made it clear that they really, really won’t like it if these cable locations are released.

Declassification studies

When the ARRB reviewed the information which agencies wished to redact from their records, they set very strict standards, demanding detailed explanations of how release would constitute “identifiable harm to the military defense, intelligence operations, law enforcement, or the conduct of foreign relations that is of such gravity that it outweighs the public interest in disclosure.” This was the standard set in the JFK Act.

As a result, CIA produced lengthy, detailed briefings explaining why they didn’t want to release sources and methods. These briefings are now themselves slated for release, and are mostly collected in the 104-10332 record set. Note that the ARRB cables discussed above are in many cases actually quite lightly redacted, with only a few items consistently held back. The declassification briefings are a different matter and some have multiple pages withheld, not just individual words or sentences.

Some of the issues in these briefings actually seem to have faded. CIA strongly argued for withholding the name of its Warrenton, Virginia records archive, but the 2017-2018 releases opened a number of references to Warrenton (though there are still places where it is still, inconsistently, withheld).

There are several lengthy briefings on why station locations should not be released, dealing specifically with demarches made by the government of [redacted] on the issue. These briefings will probably stand or fall, released or withheld, depending on how the location issue is handled.

Liaison relationships are of course closely tied to station locations. These are the subject of several briefings. Such relationships are very, very sensitive and the odds are that some of this material will not be lightly released.

Correspondence, etc.

There are good chunks of the ARRB and JFK Act correspondence that have redactions. Many of these documents are memos and letters from “focal point” people appointed inside the CIA to handle ARRB requests and briefings. Some of these liaison people were themselves undercover, and the redactions here are simply their names (their positions are mostly not redacted). I’ve written elsewhere of this issue. Releasing the names of these people, in records less than 25 years old, is bound to be another controversial decision, and I expect the CIA to argue strongly for continuing to redact these.


For me, the 1990s material is central to my interests, which center around declassification issues. I would certainly like to see as much opened as possible, but I am sure that if the release of anything in the collection meets the criterion of causing an “identifiable harm”, some of it is certainly found in the materials here.