IMPLEMENTATION OF APEX SUBCOMPARTMENTS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82M00591R000200090029-1
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 6
• Document Creation Date: December 16, 2016
• Document Release Date: October 26, 2004
• Sequence Number: 29
• Publication Date: March 5, 1980
• Content Type: MF

Approved For Releame 2004/12/22 : CIA-RDP82M00591 R00ap0090029-1
DIRECTOR OF CENTRAL INTELLIGENCE
Security Committee
SECOM-D- 126
March 1980
MEMORANDUM FOR: Special Assistant to the DCI
for Compartmentat:i_on
FROM:
STATINTL
Implementation of APEX Subcompartments
1. There have been several questions raised about
the concept of subcompartments in the APEX System and how
they would be implemented. The Navy member of the APEX
Steering Group recently posed several questions in this
regard. You have agreed to meet with his Delement STAT
for further discussion and asked me to accompany you.
This memoraudum sets forth for your review my perceptions
of subcompa.rtments, what they are and how they are to
work. If these views are different from yours, or if they
fail to take into consideration aspects of the issue that
have come to your attention, perhaps we should discuss them
prior to meeting with Navy.
2. The concept of subcompartments arose in force
recognition of the need for a select group of individuals
to have access to some information currently grouped under
the generic heading of "operational data."
3. I say "forced recognition' for several reasons.
First, the goals of APEX included the separation from all
product intelligence of details of sources and methods.
This included the traces of control, the labels of controls
as well as substantive source data. It was recognized
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1
 Approved For Releaw 2004/12/22 : CIA-RDP82M00591 R00Qp0090029-1
that the sticky tentacles of the Control System, STATINTL
the SI System and the 0 Contro ystem continued to
show on many finished intelligence products. This
restricted what was viewed as a desirable feature of a
new system - i.e., broader dissemination. of product.
Another aspect of the desire to separate source from
product arose from the recognition that the imposition. by
the collectors of their controls over product had the
undesirable results of cutting too many people in on our
operational details. And there was no gradation of this
access.
4. In our (NFIB Working Group) review of what was
operational data and what was product, it rapidly became
evident that the collectors had extended the definition
of "operational data" beyond reasonable and prudent grounds
and had in consequence caused resentment in a significant
segment of the consumer population who pushed hard for
redress. Second, there is a logical progression of
actions leading to the publication of finished intel-
ligence from collection (which includes a legitimately
wide range of activities), through processing of raw
data to intelligible form, through analytical evaluation
(including assessment of reliability and timeliness as
well as substantive content), to publication, and finally
dissemination to consumers (who require some measure of
assurance of reliability of source data - but a little
less than analysts). Unfortunately, this progession in
action does not lend itself to neat, clean defined steps.
'bhere are overlaps and gray areas that make the best
efforts at division very difficult at the least. So,
it is easier to hypothesize a nice, neat break and simpler
t.o say, "Let's do it," than it is to accomplish it.
5. In this connection we again see a reflection of
provincialism on the part of the collectors. They perceive
their collected material not as raw data with which some-
thing must be done by somebody else on a team before it
becomes finished intelligence, but rather as an end in
itself. This perception puts the processors, analysts,
resource managers, and even tasking officers on a separate
team. And, as a condition of allowing others to play
with their material, the collectors insisted that their
control rules be followed. Even further, the collectors
exercised, by claimed rightthe authority to grant the
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1
 Approved For ReleQ&e 2004/12/22: CIA-RDP82M00591 ROGQ 00090029-1
license to players. They insisted on an overview
determination on need-to-know. A considerable animus
in the consumer population has fostered over the years
because of this.
6. Acknowledging only reluctantly the essential
requirement to do something with the collected data before
it could serve the Community in any meaningful fashion,
the collectors prepared project manuals. These were
originally intended for a small group of photo interpreters
or COMINT analysts. Shortly afterwards, however, the
COMINT collectors recalled all such material. The
Imagery collectors on the other hand, seeing advertising
advantage or for whatever reason, elected to
bigger and better project manuals
They allowed these manuals to be c
broader based TK system. We now find the environment of
the SCI world in a dichotomous situation - no or too
little information available to those who need it, or
too much operational material easily available to those
who don't. I believe this condition disturbed both
extremes of the population and contributed to the pres-
sures and the ultimate review that led to APEX.
7. In the attempt to arrive at a solution, the
NFIB Working Group reviewed a large volume of product
material and identified considerable amounts that flat
out did not warrant compartmented protection; identified
a volume that was questionable; and identified another
volume that was evidently compartmented because it had
"information about" operations and another volume that
was accepted as operational. There wasn't much of an
effort to segregate these last two categories - perhaps
this should be given some additional attention by the
Steering Group.
8. The review process recognized the need for
access to "some" but not all operational data by a select
group doing processing and analytical work, and a different
portion needed by tasking officers and resource managers.
There was no attempt in the course of the review to
identify with any precision either the size or nature
of the population(s) involved or the categories of info-
mation that would he needed.
9. Attention was given to the continued and con-
tinuing claim of program managers that they alone control
release of data about their programs. This claim was not
25X1
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1
 Approved For Relee 2004/12/22 : CIA-RDP82M00591 R0GQO0090029-1
openly challenged but there was considerable sentiment
expressed in the course of long debates that such a
closed position did not acknowledge that SIOs were in
a position better to determine the need-to-know of
their populations. It became evident that a compromise
position. was essential if the goals of the exercise
were to be accomplished.
10. The form of the compromise was the concept of
operational subcompart.ments. Operational subcompartments
were to encompass that volume of information which would
be identified in-joint discussions between representatives of
SIOs involved in processing, analyzing and preparing
finished intelligence, products; representatives of the
Community elements associated with tasking and resource
management; and representatives of the national collection
programs. The requirements of the analysts, processors,
tasking officers and resource managers would be clearly
stated to the program managers. These presentations are
not to be made as a supplication or petition but are to
be accepted as good grace statements of processor/pro-
ducer needs.
11. Program managers are to receive these needs
in the same, spirit of good grace neither challenging the
appropriateness of the request nor questioning the pre-
sumption of need-to-know. Program managers are to make
every effort to provide full and satisfactory responses
to the requirements. They are not to use the opportunity
to prepare advertising brochures or to gratuitously broaden
the scope of needed information.
12. The form of the response would be a paper - a
so called operational. subcompartment manual. This product
would be presented to the requestor by the program manager.
At this point, control. over the information - the operational
subcompartment manual - passes from the program manager
to the requesting SIO.
13. It then becomes the responsibility of the
receiving SIO to exercise good faith application of need-
to-know in distributing the operational subcompartment
manual. The manual will provide the SIO with information
about some aspects of the operational. project. Through
careful examination of its contents, the SIO will find
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1
 Approved For Relea200512/22 : CIA-RDP82M00591R00Q&P0090029-1
guidance to determine who needs the information to do
his job. Those individuals will be provided with the
manual. They will he identified to the Central Registry
as authorized access to the operational subcompartment.
The identification process is the responsibility of the
SIO.
14. There was no intent, nor even discussion of the
idea, that the SIO would make available to program managers
for whatever reason, the identities of or a count of
personnel to whom the subcompartment manual was given
and for whom access was authorized.
15. This process of having the program manager or
director prepare the operational subcompartment manual
with the full participation of the "user" (in a very narrow
context) Community was viewed as serving the dichotomous
(if not antagonistic) requirements of both elements.
The "user" (noncollector) element would have full
opportunity to express his needs and have them satisfied
without hassle within the bounds of common sense and
good faith. The program manager would be in the position
of not having to provide anything unless he was asked to
do so, and then to provide only that information minimally
required to_ satisfy the request.
16. Both sides faced the need to compromise on
currently held positions. The user side has to quantify
and qualify its requests. The collectors have to provide
and relinquish control over some material. Both sides
benefit. The user receives what he needs to know. The
collector has a prime opportunity to limit the scope of
release of operational project data. The Intelligence
Community benefits as a whole by providing the mechanism
to further separate source data from product data, quell
the growing dispute over jurisdictional control of infor-
mation and satisfy the needs of other members of the
intelligence team who are required to use the information
in processing data toward finished intelligence.
17. That was the way the operational subcompartment
concept was envisioned and that was the intent of para-
graph number 7 on page 12 of the Final Report of the NI7IB
Working Group on Compartmentation. It was recognized that
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1
 Approved For Release 2004/12/22 : CIA-RDP82M00591 R00UQ 0090029-1
the concept does not equate on a one-to-one basis with the
way we do business now. But the reason. for APEX was
because the old way failed to provide satisfaction to
perceived needs.
STATINTL
Distribution:
Orig - Addressee
bhuUM (1) - Chrono
1 - SECOM Subject
SECOM fh (3/5/80)
Approved For Release 2004/12/22 : CIA-RDP82M00591 R000200090029-1