PANEL FOR CIA COMPUTER STUDY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP87M00539R000901160022-8
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 11
• Document Creation Date: December 27, 2016
• Document Release Date: August 17, 2009
• Sequence Number: 22
• Publication Date: September 26, 1984
• Content Type: REPORT

STAT
STAT
STAT
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8 j3~
-S(P 14 11
ROUTING AND
RECORD SHEET
SUBJECT: (Optional)
Panel for_ CIA Computer Study
FROM?
Harry E.'Fitzwater I
EXTENSION
NO.
DDA 84-2898/1
i
on
Deputy Director for Administrat
DATE
7D 24 Hqs
26 September 1984
TO: (Officer designation, room number, and
building)
DATE
OFFICER'S
COMMENTS (Number each comment to show from whom
RECEIVED
FORWARDED
INITIALS
to whom. Draw a line across column after each comment.)
1.
Acting DCI
7D 6011 Hqs
John,
2.
The attached represents my best
effort at drafting a guidance paper
for the proposed computer panel. I
3
have coordinated it with the ExDir-
he concurs.
4.
has agreed to chair th
pane -L.
s.
Does this meet with your approva
6.
7.
Attachment:
As stated
8.
9.
10.
11.
DDA/HEFitzwater:kmg(26Sept84)
Distribution:
Orig PRS - ADCI w/Att (by hand)
12.
1 - ER w/cy Att (by hand)
1 - DDA Subj w/cy Att
1 - DDA Chrono w/cy Att
13.
1 - HEF Chrono w/cy Att
14.
15.
FORM 61 O USE PREVIOUS
I-79 EDITIONS
GPO 1983 O o 411-632
1010
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
I. Objective. The Panel's principal objective is to determine the
most feasible and cost-effective method to prevent the continuing
displacement of personnel in the Headquarters Buildings by machines.
This includes the new building currently planned for-occupancy in the
summer of 1987.
II. Deliverables. At least three options are required. The Panel is
urged to use its own judgment on option selections. Possible options
include but not limited to the following:
A. Move all computers and storage devices to an outlying
building. This could include construction or leasing of a building
either new or renovated.
B. Contracting for provision of computer services.
C. Management system that controls growth of applications in
consonance with shrinking technology (do not let equipment outgrow
the floor space).
D. Management system that controls the data stored (and hence
the space devoted to storage).
III. Considerations. In accomplishing this study, the Panel should
consider the following in its study for options A and B:
A. Security - The impact risk and costs of providing (1)
physical protection for a remote site and its communication network,
(2) meeting COMSEC requirements for the facility and its communi-
cation, and (3) in isolating data bases from the rest of the Agency.
B. Communications - The impact on resources (costs) and
technical feasibility of communicating at the speeds and reliability
required by the Agency.
r,----,- ------- - - Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
C. Size - The study shall be based on a facility that will have
the capacity to absorb growth of machines and necessary supporting
? personnel to meet Agency computing needs through the year 2000.
D. Cost - The Panel is requested to provide budgetary estimates.
IV. Support. Staff support will be provided in the following
disciplines:
- Panel Executive Secretary (Dean Brown, DDA)
- Logistics
- Communications
- Data Processing
V. Panel Membership:
STAT
VI. Report Date:
A. Preliminary Report - 1 December 1984
B. Final Report - 15 December 1984
f7 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 ROUTING AND RECORD SHEET
STAT
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
TO: (Officer designation, room number, and
building)
1' DDCIIt,."
3' DCI
a 9q -&-74A
DATE
19 September 84
FL~~ ca~~ Po~~
T (4t- f}-Jt7A -GG o
,RY
y7 LC-fib.
A-rti G
7, L am' /-,(og-
T AT Q0C"'"e ' O(~Se
c.. ,4 P -~--re- '
Tff* -' ~.. o G l~ l ~ G ? C41- S~
APL
Tax-t ":. Y -rt+p- 1-t
Y
~ k~~'t'~r'S~ `i.?.r;''Cy.r,.;,es-..i,. ,,:e_.,a. ,..,.... .._a.. _,... _ s..,~_ --- - ----- - - ...:.74 ems, , ? ._,
}iP Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
19 September 1984
MEMORANDUM FOR: Director of Central Intelligence
VIA: Deputy Director of Central Intelligence
FROM: Executive Director,
SUBJECT: Computerization
1. Recently you sent me a note on computerization. This is
a subject of intense interest to me, not because I'm especially
interested in computers but because the future quality of our
work will increasingly be tied to our success in harnessing the
computer. What follows is a short summary of my present thinking
on the issues you raised.
2. Senior management concern about this issue isn't new; we
have even made some progress over the years in the way we bring
computer-related issues to senior levels for resolution, in
educating our senior managers about these issues (although it is
often the case that as soon as someone gets educated he leaves),
and in encouraging information sharing among CIA components. We
have also had some tentative, but obviously not decisive,
discussions of alternative approaches to the management of
information systems.
3. One approach to thinking about this problem was the
result of some strategizing between Jack Blake, then DDA, and
myself when I was Comptroller; this eventually resulted in the
hiring of an information architect. After approximately a year's'
work, the position of Information Architect was abolished and
replaced by an Information Systems Board to deal with many of the
same questions.
4. Despite this and other similar kinds of experiences over
the years, there has never developed a senior management
concensus about what ought to be done about the information
systems management issue. Indeed some people question whether
anything ought to be done at senior management levels other than
support what is conceived at lower levels.
5. Over the years, much of the hassle has been caused b.y~_
n
i
cha t h 1
g
ng ec no ogy. First, there was the mainframe mentality
pushed by IBM, which insisted that large data processing
departments full of white-coated specialists make all the
decisions. The users of such systems played second fiddle.
Indeed, the device given the user for communicating to the
mainframe system--the dumb terminal--said it all. Then came the
SECRET CL BY Signer
DECL OADR
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
miniturization of electronics, and a new role for users who were
also becoming increasingly sophisticated as they sought to deal
with technical data processing specialists. Minicomputers were
used to act as "friendly interfaces" to large mainframe systems,
helping users to use systems only data processing specialists had
previously been able to understand. Then even the minicomputers
got smaller and became microprocessors--small computers as
powerful as the older mainframes but now imbedded in "intelligent
terminals" and "personal computers" (PCs). Now the terminals
themselves can help users use complex mainframe systems and, in
addition, can do some data processing, i.e., simple analysis,
word processing, and graphics.
6. Because these smart terminals, minicomputer based
systems, and the old standby mainframe based systems are used for
different purposes, they contain different parts of our
collective information base and are used by different groups in
the organization. Over time these different groups have found
they need to communicate with one another and share their
information. Thus, networking emerged--a way of letting all the
different users of all the different systems talk to one another
and share data. We would be in clover except for two tiny
problems: the technological complexity of networking, and the
relative security issues. Networking everybody's system and data
together is not good security. Compartmentation is virtually
impossible and a really smart computer "hacker" can get his hands
on anything he wants to.
7. We and industry continue the search for compatible
hardware and software to make networking easier without denying
users capabilities they believe they need. No one vendor builds
a system that will do everything for everybody, and talk to all
the other pieces of equipment that are around. Everybody is
coping in different ways. We have coped more successfully than
many, but not without much gnashing of teeth.
8. In the course of trying to explain to you what I was
doing on this topic some weeks ago, you expressed the view that"
maybe what was required was a "point of view" about the role of
computers in our profession. I agree. An agreed set of
objectives, almost a philosophical statement of where we want to
go, is essential. Such a statement might include points like the
following:
- We. need to bring artificial intelligence techniques to
bear as rapidly as feasible on our problems, with priority
(for example) to helping the DI better integrate
intellience information and the DO recruit agents
- We need to make decisions about who will have access to
new databases before procurement decisions are made, so
that databases can be designed to meet everyone's needs in
the first instance. Security, customer needs, and other
related issues must be systematically considered.
SECRET
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
SECRET
related issues must he systematically considered.
-
It
the
in
our
is important to equip smart ADP-wise
computer-based tools they will need
the future, both to get our work done
ability to hire and retain the kinds
new employees with
to do their jobs
and to enhance
of people we
need.
- We need to limit the number of different vendors who will
develop or build equipment in our facilities, both to
reduce costs and improve security.
- We want to hold components to some database size
constraints in order to slow today's cycle in which we
design new systems, move people out of our building to
accommodate the new systems, and then try to bring the
people back. (The storage devices which hold computer-
based information occupy nearly one-half of all the
physical space we devote to computers.
9. In addition to a set of objectives, or as part of it, we
need a constitution, something which outlines what we expect to
occur at the component level, what is properly done at a
directorate level, and what should be considered centrally. Some
of the issues for such a document include:
at what level in the organization we should encourage a
serious planning effort;
- who can conduct R&D on artificial intelligence
applications;
who will manage certain services of common concern;
- which components are competent to oversee the development
by new contractors of future capabilities and which should
be required to use a service of common concern.
10. Finally, we need a sensible process which brings
certain information systems management issues to senior
management attention. Now, senior management review of our
programs mostly occur as part of the budget process. One can
quarrel with how well we deal with these decisions but, in
principle, the budget process is a sensible place for priorities
to be sorted. It is not an adequate vehicle, however, for a
discussion of our overall objectives or for consideration of
issues of compatibility between systems.
11. The steps enumerated above are important, but just as
important as taking a series of procedural steps is finding a way
to exert some discreet 7th floor leadership on these issues. We
need to involve the DD's even more directly in this problem than
we have to date. You recently expressed interest in the AIM
SECRET
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
system. Bob and Evan are using it, Harrv has a terminal though I
don't believe uses it very often, and Clair isn't even this far
along (yet)! I have just taken steps, with ODP's help, to link
together electronically certain of our senior people--including
the four deputies, and Chuck Briggs. This is a start
toward involving more of our senior people in the use of EDP
tools. I am confident that with this involvement will come
pressures downward for compatibility and for access to databases;
in short, leadership.
12. As a footnote, we have a few advocates of
organizational change to enable us to.deal better with
information management issues. Some favor creation of a fifth
directorate--generally conceived as an amalgam of the Office of
Data Processing and the Office of Communications. This approach,
however, I think puts the cart before the horse. It seems to me
that significant organizational change should follow conscious
decisions about what we want to do, not the reverse.
13. Taking advantage of our past experience, I am working
to pull together proposals in each of the areas mentioned above
for further consideration and action. I hope to have this done
in relatively short order. You quite rightly stressed in your
note the need for senior management involvement. I am equally
anxious to avoid the mistakes I and others have made in dealing
with these issues in the past. I am confident that if we can
engage the Deputy Directors at the right level, a level which
matches their understanding of their organizations. and their
problems, we can successfully fashion effective tools to help us
deal with the difficult management issues we will continue to
face as computers become an ever more integral part of our
profession.
James H. T ylor
ORIG: ExDir/JT:smg (9-19-84)
Distribution:
Orig - addse
1 - ExDir
1 - ER--,a
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 135Lv 984
4
25X1
25X1
NOTE FOR: Associate Deputy Director for Intelligence
Associate Deputy Director for Operations
SUBJECT: Computer-Assisted Dissemination
of Sensitive Intelligence
I would like to meet briefly with you to discuss the DCI's
suggestion to the DO for a computer system to disseminate
selected sensitive intelligence reports to senior policymakers.
Before we ask our ADP, communications, and security experts what
is possible, I think we need to agree among ourselves on what, if
anything, we want to do and then proceed to do it together. Some
of the issues we need to resolve include:
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
C()JVrJL)f f I A i Elt 84-9020
? What intelligence should the system handle --
"Exclusive For," individual NID and PDB articles, "blue
stripes?"
? How many senior customers will it serve and where are
they located? Should they have terminals only in their
offices? Will portable terminals with encrypted
communications be necessary?
? What capabilities will the system offer -- text only or
the full range of graphics, maps, and photographs? Will
customers be able to ask questions?
? Will password protection of the system be adequate
(some customers might be tempted to give their passwords to
assistants or leave the terminal unattended once logged on)
or will some more advanced access mechanism be needed?
? How much are we willing to spend for such a system?
When do we want to make it available to our customers?
My secretary will arrange an early, mutually convenient
meeting. I want to get a quick sense of where we should be going
and start the technical experts working as soon as possible. F
James H. Taylor
Distribution:
1 - ADDI
1 - ADDO
ExDir
1
_P 1.
HD r~-tIVY/.I Al. 13 SEP 1984 CO,il C! ) ;41'!"
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
DDCI?''_
Chm/NIC -
DATE
AO/DCI
C/IPD/OIS
:J'. Ri xse
Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
 Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8
Executive i 1~ jictrTf
84., 807 l ..~.~1
6 September 1984
MEMORANDUM FOR: Executive Director
FROM: Director of Central Intelligence
SUBJECT: Computerization
1. Some weeks ago, I was concerned with the problems we were having
with the breakdown of the WANGs to the extent that I called Dr. Wang to
have him straighten it out and he sent a team out to look into it. I am
concerned that there are more fundamental problems in the way we have been
approaching computerization, particularly with each Directorate selecting
its own system. This results i a DI analyst being unable to communicate
through his terminal with a DO reports officer, data bases which cannot be
used interchangeably with different terminals and in connection with one
another, inability to use Community data bases on-line in other than the
originating agencies, etc.
2. I'd like to understand the role of the information handling
architect, the Information Handling Committee, the data processing unit
in DIA, and their respective effectiveness. Above all, where and how should
we be going from here. I'd like to work out a comprehensive and speedy
method of assessing this and charting directions. As a first step, we should
make sure that senior management is on top of this problem and prepared to
develop a coordinated strategy and head the Agency as a whole in a new
direction, while taking into account the specialized needs of each component.
3. One reason for addressing this promptly is the large investment that
the new building will require in additional and relocated communications and
computer commitments.
William J. Casey
T__ _ Approved For Release 2009/08/17: CIA-RDP87M00539R000901160022-8