MANUSCRIPT REVIEW

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP85B00236R000200020004-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 42
• Document Creation Date: December 12, 2016
• Document Release Date: March 29, 2001
• Sequence Number: 4
• Publication Date: April 28, 1980
• Content Type: MF

CONF ID
?Approved For Wease 2002/01/08 : CIA- DP85EO0236Q000200020004-7
28 April 1980
OGC Has Reviewed
MEMORANDUM FOR: All. CAD Personnel
25X1A FROM
Chief, Classification Review Division
SUBJECT Manuscript Review
1. CRl) recently reviewed a particularly troublesome manuscript submitted
by a former employee. Several days of meetings ensued, during which the recom-
mendations made concerning the manuscript by the Agency components Involved were
examined in exhaustive detail, reconciled where feasible, and in many instances
abandoned. Representatives of the OCC (which had requested this decidedly unusual
procedure), the DO, and the Agency Publications Review Board (PRB) participated
in these sessions, as did the CRI) reviewer (on behalf of the DA). (U)
2. This case was complicated by a lawsuit which the manuscript's author
had previously filed to enjoin the Agency from "censoring" his book. Only after
the Supreme Court decided the Snapp case in the Agency's favor did this author
submit his own manuscript for review--preparing at the same time to present legal
challenges to any and all deletions or revisions the Agency might demand. Hence
OCC's participation in this particular review exercise. (U)
3. 0GC takes the position that revision of manuscripts submitted for review
should be confined to material which is demonstrably classified or classifiable
under Ex?.c_utive Order 12065 and/or such legislation as the Atomic Energy Act of
1954. The National Security Act of 1947 and the CIA Act of 1949 exempt the Agency
from making certain disclosua:es but, according to OCC, these exemptions are not
applicable to information from former Agency employees. The categories of infor-
mation defined in the two Acts were, in general terms, "grandfathered" into E.O.
12065 under. Section 1-301(c) -- ":intelligence activities, sources or methods" --
and are covered in considerable detail under the corresponding portion (paragraph
9cc) of our own specific classification criteria Agency Information
25X1A Security Program Handbook). (U)
4. However, such criteria do not in themselves validate the classification
of information. We must also determine whether there exists a reasonable expec-
tation of "at least identifiable damage to the national security" (i.e., per E.O.
12055, to the "defense and foreign relations of the United States") if the infor-
mation is disclosed. No identifiable damage, no classification or classificability.
The Order (Section 1-3U3) and Handbook (paragraph 5c) make a rp esum tion of such
damage as regards disclosure of "foreign government information" (which includes
25X1A liaison information and is further defined in- paragraph 9b) or of "the
identity of a confidential foreign source" (interpretable as including, for example,
a liaison service but usually meaning an individual, "warm body" source of HUMINT).
(U)
Approved For FRei'e*4 2002/91/08: IA-RDP85B00238R006260'0~d0043tc'-4..-Z---- ------
 Approved For fir ease 2002/01/08 : CIA-RDP85B0023OW00200020004-7
5. According to OGC, the above presumptions can be .invalidated by a showing
that the information at issue is already in the public domain or even, perhaps,
that the foreign government or other ""conffdenti.sl :source" will not be affected
by its disclosure. Here the "burden of proof" would seem to be on the author to
make, the case against any "identifiable damage". but once litigation has begun
almost anything can happen. (U)
6. In the Marchetti case of some years back, the Agency wat; upheld by the
trial court and Marchetti's book was published with blank spaces representing
the deletions CIA had demanded. Marchetti appealed, however, and the appeals
tribunal restored a number of the originally deleted items -- whereupon Marchetti
published a new edition with those items printed in bold-face type. They were
thus called emphatically to the reader's attention, making the situation worse
from the Agency's standpoint than if those items had not been deleted in the
first place. (U)
7. When a manuscript becomes an actual (or even potential) matter for
litigation, authors can and will wrangle over every paragraph and sentence,
even over individual words and phrases. It does the Agency little good, and
indeed some harm in terms of wasted time and effort, to mark whole chapters
and pages for deletion as did some of the reviewers (not CRD's) who processed
the manuscript cited in the first paragraph above. Only rarely can the dele-
tion of an entire paragraph be justified in terms of proving to a court of law
that all of that paragraph's content is classifiable. Similarly, general
statements about the classified nature of material in a manuscript are useful
only to the extent that they may tell some other reviewing unit what to look
for. That unit must then make the actual deletions (or other recommendations
for removal of classified material) which the PRIB can present to the author. (U)
25X1A 8. As noted (see attachment to tloist memo), the PRB has only 30
days from the date a manuscript comes, in to return it to its author with all
required revisions indicated. It is best, therefore, for the initial manuscript
reviewers to do as complete a job as possible. CRD has the responsibility to
review "for the DA with the. exception of the Office of Security" and is thus
(`rith this one exception) the final as well as the initial reviewing component
for the D. Parenthetically, the DO works differently. A manuscript may he
sent to several Area Divisions, Staffs, etc. before DO/IMS conducts the final
Des review based on the reconnuendations of other units. Of course, in our review
we can and do call to the attention of other Agency components material which
appears to be of interest to them. (U)
9. Unlike CRD's basic systematic review function, manuscript review is
sscsentially a sanitization procedure which somewhat resembles our routine
processing of Foreign Relations of the United States (FRUS) submissions from
the State Department and the occasional Freedom of Information Act (FOIA)
ci,ses assigned to us. It differs from these, however, in that what appears
it, an employee's or ex-employee's unofficially published work does not
constitute an official Agency disclosure. Agency approval for publication in
no way endorsee or validates the published material. (See paragraph 2b(9) of
25X1A - although authors are not, of course, required to use the disclaimer
which this provision of the regulation "encourages.") (U)
Approved For Release, ?002/q1tg8 : CIA-RDP85B00236R000200020004-7
s a, ., , a
 6
Approved For lWea's~ IN /0 08 : (tIA dP86$0023 900200020004-7
10. Accounts of positive intelligence, Cl or CA operations as well as of
other Agency activities in which employees or former employees wore involved,
or learned about during their Agency careers, frequently do contain classified
or classifiable information even when the authors make some effort on their own
to disguise or obfuscate the true facts. Even when manuscripts are billed as
fiction the authors often fail, in our judgment, to conceal actual circumstances
adequately. This can lead to identification of real intelligence sources and
methods, or expose sensitive liaison relationships. It was William Colby's
disclosures in his book Honorable Men (which was presented as fact) that led
25X1A to and the establishment of the present manuscript review procedures.
The evRe of Honorable Men under the system tormerly in effect failed to provide
for adequate sanitiration. (U)
11. Most ostensibly fictional works presented for review are cast in the
traditional "spy story" mold pioneered by such authors as E. Phillips pppenheim,
Somerset Maughom (who once served as a British intelligence officer), Graham
Greene, and numerous more recent exponents of the genre like Howard Hunt and
even William F. Buckley. The degree to which Agency authors, whether present
or former employees, preserve security by disguising actual people, places,
and events in their stories obviously varies widely. In dealing with fictional
works the reviewer must try to assess the extent to which the author's actual
experience may be mirrored in the stories he tells. The reviewer must assumes
that authors will be identified in the book's jacket blurb or otherwise as "CIA
operatives" or the like. Some details of their careers may also be published.
Any such identification will. clearly enhance an author's credibility and make
the work seem authentic if not official and authoritative, requiring closer
screening by us. If necessary, the reviewer can obtain information about the
author's career in the Agency. (U)
12. Non-fiction about the Ag;vnr.y by former employees is normally published
with mention of the author's CIA connection. Most manuscripts in this category
are biographical -- usually autobiographical. Authors are discussing their own
careers and grinding their own particular axes (some authors of fiction do this
also, to be sure). Again it would prove useful to reviewers of non-fiction to
be aware not only of the basic facts about the authors' careers but also, in the
rase of farmer employees, of tILOtr cover situation on leaving the Agency and what
they were permitted to put in the unclassified job resumes and other material
prepared for each departing employee. Review actions filtered through such basic
Information should be more realistic and supportable, thus providing a more pro-
fesssional review position. (U)
13. It clearly makes no sense to demand the deletion from manuscripts of
information the authors have bcen allowed to disclose, or which has been disclosed
on their behalf by the Agency itself (e.g., by the retirement elements of the
Office of Personnel). Certain proposed revisions to the manuscript discussed in
paragraph 1 above were prevented by such disclosures in that author's particular
case. The nature of some of the work this author did overseas, moreover, had
been disclosed by an earlier author who was also a former Agency employee. The
fact that these data were already in the public domain prevented other proposed
deletions from the subject manuscript. (U)
Approved For Rele(( 4qQ?40'1J G1A' Dr 85BOO236R000200020004-7
4.oi_~ JJ 4
 0'r N F~ DF',1J1T,j'
Approved Fori lease 2002/01/08 : CIA-RDP85B0023SWO0200020004-7
14. In the interests of saving time and effort, givin,; the PRB the best
possible DA reviews of manuscripts, and thus generally ensuring maintenance of
the same high standards in this area of our assigned responsibilities as in
others, CRD's manuscript reviewers should:
a. Read through each manuscript at least once before marking anything
in it for revision, to see what the author is about and to avoid messing
up the manuscript only to find that a seemingly necessary deletion on page
27 is invalidated by material on page 532 (or even in an appendix, etc.).
It is, of course, desirable to paperclip pages containing any apparently
classified or classifiable items spotted during this initial reading, or
to take separate notes thereon during the process.
b. Go through the manuscript again and mark on it the items which meet
classification criteria (paragraphs 4-6 above). This will not normally
involve whole pages or even paragraphs (though there will be occasional
exceptions to this general principle). Rather, only those sentences phrases
or words which constitute: the actual basis for classification should be
marked. If ---- as will often happen -- this results in portions of the manu-
script looking like the proverbial "Swiss cheese," so be it. It is up to
the author to determine whether to "edit around" Agency-required deletions
or to publish a text. full of holes, a la Marchetti. NOTE: It is best to
mark such items with a "yellow highlighter" which will not reproduce on
copying machines. Errors made during the marking process can then easily
be corrected: simply copy the relevant pages, make the corrections and
substitute correct for Incorrect pages. Manuscripts so marked look much
"cleaner" than is the case when bracketing or underlining is used. To make
bulky manuscripts or those requiring a great deal of marking-up easier to
handle, they may be put into loose-leaf binders while being reviewed.
Although this may involve a little extra work to punch the pages (they are
sometimes already punched), it facilitates page-flipping and avoids making
CRD look sloppy -- as we sometimes do when what goes back to PRE (or even
to the author) is tattered or dog-eared. It is also useful to number the
lines of text on manuscript pages marked for two or more revisions, which
can then be easily identified when the review results are written up.)
L Read the manuscript once more, for quality control. Make sure there
is a valid classification basis for each deletion or other revision indicated.
Such rechecking will also often pick up items missed earlier, or show that
some previously marked revisions are actually unnecessary or should be
amended.
d. Finally, write up the review results. t?ach deletion should be speci-
fically identified. When several fall within a clearly definable and clas-
sifiable area (e.g., "information which could identify a covert agent" or
the like), they may all be grouped under a statement to that effect. Do not,
however, use such overly broad statements as "intelligence sources/methods/
activities" without pecifying how the item selected for deletion relates
substantively to the E.O. 12065 classification category. Most of the items
we mark for deletion will be selected on the basis of intelligence sources,
methods and activities for one reason or another; merely to "parrot" the
classification category will convey no useful information to the PRB. It
-4-
Approved For Release r2002101 %0f, CIA-RDP85BO0236R000200020004-7
 Approved For Iase 2002/01/08 CIA-RDP85B00236iD0200020004-7
should also be kept. in mind that one type of classifiable data often found
in manuscripts -- liaison relationships and activities -- fits better
under the "foreign government information" provision of E.O. 12065 (Section
1-301b) than under the "...sources, methods..." provision (1-301c). The
latter is sometimes over-u,;ed to cover anything and everything, when one of
the others could more fittingLy be cited .in many instances. The "identifiable
damage" criterion applies to all classification categories --- see paragraph 4
above. No deletion should be identified unless the reviewer can "reaao.cably
expect" identifiable damage to occur if the manuscript is published with the
original text unchanged. (U)
15. When reviewing mrinusc.ripts written by former Agency employees:
Release: Mention of the Existence of Agency Stations in countries *,ere
the author was assigned PCS under official or no cover. When personnel who served
overseas under official cover ere separated from the Agency in overt status,
"opened up" as the phrase goes, it is clearly impractical to try to conceal the
fact that they worked at Agency Stnitiona during periods when they were on official
tours of duty abroad. It is, of course, essential to ascertain whether the author's
career was "opened up" upun bel-aration from the Agency.
Delete: The .i entity cF Stations where the author served only TDY or at
25X1A which he served or under non--official cover. Also delete information
which identifies other Agency t.?erse,nnel serving at any Station, together with
details of specific operations or ether activities at any Station and data on any
Station functions, organization, numbers of personnel, facilities, methods of
operating, etc.
Release: The mention of Headquarters organizational units at or above
thq Office level (Division level, in the.DO) and also below that level when
eneric designations are used (e.g., ""the which does not accurately
name an actual component so designated but teensy re ers to a unit function).
Delete: An actual Aecncv desiunattnn 4ciph .. "
Release: The names of other present or former Agency personnel who have
authored previous publications and whose Agency connections were thereby disclosed.
(NOTE: l'RB is preparing a macl+ine:?-records listing of such authors, the titles of
their manuscript submissions'and publications, and other data to serve as a
'"reference shelf" for future use by manuscript reviewers.)
25X1A
Release: The name and location of the Agency's
facility. This was among the dt:;closiaes Marchetti was ultimately a owe to pub-
lish and therefore cannot be di_letuct from the manuscripts of other authors. Keep
in mind that the Marchetti. incldeiir does not constitute authorized executive
disclosure; we continue to deny in:ormal.ion about to FOIA requesters, 25X1A
and we do not declassify ouch Inform ition when encoun ere curing systematic
Approved F (n r apse 01 f
10 r/!lRDP85B00236R0002000200047
 N D: i T; A I-
- Approved For ease 2002/01/08: CIA-RDP85B00230M00200020004-7
16. Many if not most of the above "do's and don'ts," along with a number of
others, are already familiar in general terms to CRD reviewers. They are included
primarily because they concern issues which arose during the final review sessions
on the manuscript cited in paragraph 1, and say help to sharpen other reviewer's
perceptions as they did those of the CRDer who took part in that revief4. (U)
Attach tt Package under routing shsat entitled "Procedures for Submission of
Man scripts to the Publiceticas Review Board (PRB)," enloaing eeno-
random (sear title) from Exec tive Secretary of the PRB,"- 25X1A
(cited above -- emphasis lied) and Tabs A, I and C (explained
in PRB/Exec's memo.)
Approved For ReIel66!;24Q2/p;118 -:`~gIA-F DP85B00236R000200020004-7
 Approved Foirpelee 2002/01/08: CIA-RDP85B0023000 0020004-7
25 September 1980
25X1A MEMORANDUM FOR:
ITB/PM/OI S
25X1A VIA:
C/INZEL CRD/0IS
25X1A FROM: "N W~ C)
SUBJECT: Request for a Printout from the DARE System (U)
1. We have a request for a printout from the DARE System of all the
OSS entries which are entered under Job #79-00332A. It is requested that
the run be made on the basis of characters 9 through 25 of the Document
Identification Number section of the OSS 4023A form. Thus the printout will
list entries by alphabetical and numerical order according to their "Unit,"
"Branch," "Category," "File Number," and "Folder or Document Number."
See the attached copy of an OSS 4023A form for an example of the elements
"b" through "f" of the Document Identification Number section which are to
be used. Include full Subject Titles in the printout. (C)
2 Please deliver one copy of the printout to:
25X1A IMSMIN Room 1 D 27 in Headquarters, Ext. 6865. Please deliver the
secon copy to C/CRD/OPS, 211 Key Bldg., Ext. 2777. (C)
3. Note that during the early stage of input to Job #79-00332A, some
errors were made by using zeroes for the letter "0" and vice versa. Only
the letter "0" should appear in character positions 9 through 17, i.e.,
for "Unit," "Branch," and "Category," and only zeroes in character
positions 18 through 25, i.e., for "File Number," and "Folder or Document
Number." Some of these errors were corrected but it is possible that not
all of them were. If that is the case, would you also correct that situa-
tion before.making the printout requested above. (C)
Attachment: As Stated
Distribution:
Original - Addressee w/att
1 - CRD/ Intel
DEIMVATIVE CL BY _
i i DECL R 4 EVW ON
1 - CRD oved%or Releas 6ai0" ,A`=1 6 00236R0002000200( --G~
Clara -. cV......
 LLEGIB
L Approved For Release 2002/01/08 : CIA-RDP85B00236R000200020004-7
Next 3 Page(s) In Document Exempt
Approved For Release 2002/01/08 : CIA-RDP85B00236R000200020004-7
 Approved For RM?ase 2002/01/08 : CIA-RDP85B00236QO00200020004-7
Rtlt .nfvnnc' JV4N"7~~.
t1 l (Tian i t ica I, ioil:
", ~ f dJ-44 * ws , S
I'h( rarth~_~r;tr nd n,'cd f(: cl l i;~ ,.lterial is
c 1 yle t ta~na i ('n. raI i .e
~n;' 3ra~T r f ed i c S ( stern star+c]t;oint,
ce c.crnc.('pi of 04
tri f ied and/or released ;.nfor,-nn, tion.
it of these -urreire_ s ' ;t_ems were desined and built
r *. a sec. i f i c Tulrpose 'n(1 u i l net: i ndo Indent ly support
~l= 1< enc~ rric.le rec!~ai re Brit ~':t-,x~~lr :1 u r iolr rodesigr and
oprnent C't tort,
(~i3 icial release of in~orrnat on:
l f,~ rc' are many (aperox i mate ll IS) off;' es or ind-ivi.d-
:I[-; in the 1.,;ency who officially have authority to release
intonnati_o11 to the public. This ~fdd complexity to the
i~r(lh]err () f d]erltifying and record.in.c; the' releases.
Destruction of sanit i zed (loci mtent :
)(tcruneuts which ch are sanitized aiiui rcaleased to the public
c:r,rrsidered ternpor?arv records and are destroyed at the end
ire \'fars.
! tec.omrrl(~r,d a t ion :
cll., s;a'ne c'nr_l_Dne-e t he i ra' r,, a o:i - ed t~~ the- puhl is .
I ::'C]ucc the ]1( ss I hi l 1 t.+,' e l l } i er"t'nt ;rin i 3 z(scl viers ior' - of
The s;iinta inillg the interface would
he co-:t ly in tr.nn o software and hardware. A ;econd reason is the
under sta ldah lC d ss iraahi I ity on the hart of the P1)0 to restrict access
to tiicii automated Sv'Stenis. A third is that alt of the existing
'tit evils are rno' on i 1.ne Which ii-':ins an automatic access would be
:arr rnt lv imposti i_ble.
1 :f f k)N 4 I ,
cce lol> :, 1ICi4 ;stem spec i f i ca it y designed to capture, store, and
r e t . i t vu this intormation. Contrail the control and management of
'ris s,': t.eru in one componcnat. i :I1 a clear dc#inition of the types
lat"I to be c(--llected and the methods to be used for storing and
i' t FI infornn,lt.ion.
This opt ion will also require the support of upper-level Agency
rlanagr. tc~nt for a commitment of rik)ney and people, as well as Agency-
_~u i daar~ e rxla.i;. ring ;111 corxllotle=nt; to provide information to be
ror~~d in th 'vsterra.
the cos; of (icavcloplrlg su: }_ a system would depend very much on the
> cif i,rformat ic'n regrl red to of Cect ivel r accomplish the task of
rac;:in{; declassified and rc leased docr>nen-s. Experience tells us that
i1 ,h srrn;flc t indox ;y,,tern ,tiill be expensive to develop and maintain.
I f buU 1iraal:e the a,'slmipton that this option Will be built using ODP
innle ;ts and existing facilitLes such as the ON) Standard Data Base
Ian,i,,enient Systems aind r...>nninals you can get some idea of what the
iar7iinl, and development might cost. The .study would probably take one
oaur?enr ain,.l the development from six months to one year which could cost
lao!"i 190,000 to 150, 000 dollars. The yearly maintenance costs would
Jcp ncl o f `t e tvpc of system, number of reque its,, and volume of data
n1v~dyed. The I reatest initial expense would be in going back and
II)t-, ri rg all of the previously.' released infonnati_on and attain the cost
vcr?v difficult to estimate w~~ithooot S(-me idea of the volrune of data
1.rvolve.l or tixe cethod which would he used to ;ate-an>l,.Iish the task.
Approved For Release 2002/01/08 CIA-RDP85B00236R000200020004-7
 Approved For Lease 2002/01/08 : CIA-RDP85B0023OWb0200020004-7
in ori,i(_`I to 13C a i l l t . ~ to r COlllnlc`nd ~1?ll' :,l tll(.' optlurl`,, management
(I-,t ,rrswer the follownu. CIUestiQrrs:
(1) tv)rrt is the ultimate go"] t? ? TO
7 l I a. _ 1 _ n .. w !t"' fi~'~ - .F._ L. ,n6-G,.. P .~ r '~s#.',..?G-. ~d /-. C.'
-e K- c . ti.. 4-' _-s t .d--
(3) :iowv lrwrc"h rnonk?v, rlrarr;)owcr, ;e Id time is it W0301 `r`
-ter those (;uest:ions have been r3nsi;oreel, if we are still intere _;ted
ire I)UI-SUinl, one of the options, then I would recommend that you build a
new system, (OPTION 4).
(;ontinuation of current effoi ts:
Irradequ,rte and not rleetitrg our current needs.
(-)PTION (2)
MMlodi #_rcation of an ex i sting autor;rat.ed syystern:
}:~t>crionce tells lr, t"hat t}l.:s ,l}~r~roach is neither ecoromi.cal
nor ~rr~; es ;ful .
:o#tlti;rrce link:
Not technic;3l 1i feasible= at tall i s Lime and if it were it
.,ouId meet the most resi
;.'ovo1 ;lanellt t: a new ;i,rtoma.te;l
:t': 1011 lt_ an old prob'+err..
The l,ui10 nr, It a mew ~.v trm 3s mr;~ tll:lrl just developing hardware
1n.) e twlrc.1 to handle data. Me most critical decisions, especially
ul sv:;t .nr uei a~ 111J4 one, r~ ll l he:
1)a lvc !live the complete lrldcz_standsrig and concurrence
of top AfI)ency managemolt and o,F those components which
will he affected by rho new system:'
w ~
Itiit3o , i c be re pons3lNl r1 for the ~ ~rlcloprent, r&ain- D/S
t(`31amn.c r3, I dir.} C ~e'ration of the nc'w systt1T.`:
Approved For Release 2002/01/08 : CIA-RDP85B00236R000200020004-7
 Approved For I ase 2002/01/08 : CIA-RDP85B0023OW0200020004-7
?,! rsta1)1 7 1ea.rly cir(;rcc1 n}7ic'rtiv ;arid determince
%hrrt inirIII, tt inn +1)l1.! ,c~ cc71 -(-1 "d trc ;rrc~omlcl i h
i ,> ; tawo~e the most 0,.-onomi cal sy-tem available that
n 17 srItisfv? your ohiective. }'here :ire matey optiorl4
iwthi i:rhlc and hefor - c:7e can he chc,.cn with confidence,
r; r ;rn;t i i 1'0 IVIU? ,rood d i IT) i r i( n of its nurnose.
C)nipolIerit r?~~po iE~1 1.72' cal,t ,r in r11~Onnat`Un about docIZ-ants
At 1u,-(, r)een _ll'~'(a"-1} I `d li t] r 1eati_~} iron he more thin a place
i`.1.0 ()thers send paiper h)ccau f' that i the T'Ule. In Order to make
'uch a systcni work properly, this component should have prior knowledge
of re]cc.1ses will the single releasing authority. The releasing authority
Voc.1d ho no more than a signature which indicates that the information
i: now part of th:' System and not authority to evaluate the content of
r~l0. or the individini ':; right to 'e]ease the infonr>ihtion.
f orther -ecocriulicnd that an am ilv, t t r'ow the Off ice of Data Processing
,0 ~~ i:'ne I to C,vrltu,rtC the ii(ccds And iaake an informal reeorrnendatiorp
oa t}hairs of ,'ai.htO7m:ctedi '+v-stem needed to -tore and retrieve this volume
~i i7hil)rrmit.ian.
.Z, Avg 17
Approved For Release 2002/01/08 : CIA-RDP85B00236R000200020004-7
 Approved FoorRelease 2002/01/08 : CIA-RDP85B00236J:Z000200020004-7
'I'lle three automated system.- descried helow are probably riot the
cn]' r a n tom,l t.t d P_ toms or Iii which c.t:. i n the Agency and contain
:,r,fflc 1nt'onlljltion i})Ont dec1a;sIlie'i or rek"Ised doctntent_a, but they are
~:a-:t Edit pity the major ho](Ii~~~s.
`11ae MIT system is used b,r the Cla:-;sification (review Division of
the lnlonnat:i.on ?-,ervic.e:; Staff (CRU)/ISS) to document the decisions made
n the systematic review of permanent Agency records for declassification.
tans in the near future to store iniernmation on all declassified
>_'rtr.sent records throughout the Agency as veil a all. -1assi.f_i.cation
`t ilsion deters nations recluiririg I`t'1