SEMINAR ON AUTHENTICATION OF AGENT SOURCES OF INFORMATION

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80-00630A000200010001-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 25
• Document Creation Date: December 15, 2016
• Document Release Date: September 19, 2002
• Sequence Number: 1
• Publication Date: August 9, 1976
• Content Type: REPORT

CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
SEMINAR
REPORT
CENTER FOR THE STUDY OF INTELLIGENCE
CENTRAL INTELLIGENCE AGENCY
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
THE CENTER FOR THE STUDY OF INTELLIGENCE IN OTR OPERATES
A RESEARCH AND DISCUSSION PROGRAM KEYED TO THE PROCESSES
AND FUNCTIONS OF INTELLIGENCE. THE OBJECTIVE OF THE CENTER
IS TO CONTRIBUTE TO THE PROFESSIONAL UNDERSTANDING AND
TO THE RECORD OF THE ART OF INTELLIGENCE. RESEARCH PROJECTS
ARE UNDERTAKEN BY INTELLIGENCE "FELLOWS"--VOLUNTEER
OFFICERS FROM ACROSS THE AGENCY ON FULL-TIME DETAIL TO
THE CENTER. INQUIRIES ABOUT THE CENTER PROGRAM OR COMMENTS
ON THIS REPORT ARE INVITED BY THE DIRECTOR/CSI, 25X1
25(1
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
 CONFIDENTIAL
Approved Fo Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
CENTER FOR THE STUDY OF INTELLIGENCE
9 August 1976
Seminar on Authentication of Agent Sources of Information
IS it necessary to introduce more structure, including
perhaps Directorate-wide criteria or formal guidelines into
the agent authentication process? This was the basic ques-
tion posed to some 35 intelligence officers primarily from
the DDO, but including representatives of OMS, OTS and the
Office of Security, who gathered recently in OTR for an
1
afternoon seminar on the subject of agent authentication.
The group consensus was decidedly against set criteria-
circumstances and interests in each agent case are different
and set criteria would. tend not to be an aid, but a straight-
jacket. Still, improvement in understanding and attention to
the process of authentication across the Directorate is needed,
according to the seminar participants. It should be tackled
in two ways: more focused training and more involvement in
the process by other officers in the intelligence chain
(reports officers and analysts', etc.) in addition to the
case officer.
The seminar was one in a series sponsored by the Center
for the Study of Intelligence designed to be of interest to
the Operations Directorate.._A.list of attendees at the session.
is attached to this report. An issues paper on the subject
was circulated before the seminar was held; a copy is attached
to this report.
Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
CONFTDF.NTT Ai. I 125X1
 Approved Fo1 Release 2002/11/04: CIA-RDP80-00630A000200010001-7
Training Needs
According to several seminar participants there is a
lack of attention to the principle that the authentication
process is a continuing obligation and should be a continuing
focus of concern by the collectors. One way to overcome this
would be to give more attention in training to the need to
take all the applicable steps in the process:;as a matter of
routine. As we do it now, the training for case officers
does not include instruction in any formal, systematized
authentication process as such, although trainees do get a
two-hour biock on some basic-testing procedures. One reason
for this lack of emphasis is the absence of written doctrinal
material for training that sets forth a generally applicable,
logical sequence of steps for authentication. Case histories,
including nitty-gritty details of most of the steps:; in the
process, are also lacking.
25X1A
Approved For ReleasC(al)tIRI W-RDP80-00630A000200010001-7
 CONFIDENTIAL
Approved For'Release 2002/11/04: CIA-RDP80-00630A000200010001-7
25X1A
A similar need for more training and familiarization was
voiced in connection with psychological. assessments. In
response to the criticism that the psychological examina-
tions frequently seem to result in assessment language so
hedged as to be relatively useless as an authentication tool,
the experts in the field contended that. results could be
improved if more care and preparation went into carefully
outlining the objectives for each individual case being
examined and in giving the experts proper data on the agent
so that refined and useful questions could be developed.
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 Approved Fc Release 2002/11/04: CIA-RDP80-00636A000200010001-7
25X1A
One-week courses are available to case officers on the use
of the
CONFIDENTIAL
an intelligence and personality measure-
ment system.
Apart from the functional aspects of training, there
were some attitudinal issues raised. One concerned the
dynamics of the collection process, which some participants
viewed as now favoring numbers of recruitments and numbers
of reports from these recruitments rather than quality.
The result is to build in a fear on the part of case
officers of ruining the prospects of a recruitment. and
reporting operation by being too strict about authentica-
tion procedures. This can be the opening wedge for fabri-
cations. Again, an educational effort might help reverse
this attitude on the part of case officers by assuring them
that the Agency is vitally interested and willing to reward
a serious and responsible effort to cut off marginal reporting
operations,because they will not stand strict authentication
procedures including a close examination of the value and
accuracy of the information. Indeed one participant sug-
gested our training include the criteria used in journalism
schools for proof of access and accuracy--where libel laws
make the matter a serious one.
Other Links in the Chain
Several participants saw a weakness in the authentication
process due to the lack of a central, responsible individual
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00636A000200010001-7
in each Division charged with overseeing authentication.
Some present said they believed the case officer should
be the responsible person. In the EA Division system--
the only formalized authentication process required by
any Division--the COS is designated as the responsible
person. At least two participants questioned whether there
was not a useful separation in authentication procedures to
be observed in terms of the responsibility for authenticating
a source (agent) and the authentication of his information.
But others expressed the belief that the two are really
inseparable and should not'be isolated. In denied and
semi-denied areas, where access to the agent is highly
restricted, there is naturally more dependence upon examina-
tion of the information reported than in accessible and
hospitable areas where the agent is in regular, unhurried
contact with his case officer. But it was the'general view
that an education process is'necessary in the Directorate
to help develop understanding of the need for, and the
practice of, bringing all elements into the authentication
process that properly have a role. This includes the case
officer, the Headquarters desk officers and reports officers,
the analysts,
and the psychological
assessment officers. The relationship might be expressed
25X1A
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
--the case officer: He is in most direct contact
with the agent or prospective agent. He must
be aware of the need to call in the other
players in the process of authentication and
to do his utmost to use all the tools available
that can reasonably.be applied to his agent as
part of the process. Ideally, he should prepare
each step in the most careful and meticulous
manner, and not feel inhibited from finding
negative as well as positive elements in his
case. There should be no opprobrium attached
to uncovering negative elements.
--the Headquarters desk officer: He should play
a review role in each major development con-
nected with the operation to assure that the
case officer is not overlooking significant
authentication aspects of the case and is
properly.calling upon the support available
at Headquarters.
--the reports officer: He should assure that
before an agent is "pitched" in a recruitment
attempt there is really a potential that the
agent can, or is likely to in the future, supply
valuable, needed information. All too often
this vital link is not established before
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 ~~ CONFIDENTIAL
Approved Fo-r Release 2002/11/04: CIA-RDP80-00630kA000200010001-7
recruitment, and the reports officer is asked
to supply requirements for a recruited agent
who does not have very valuable access for what
the Agency really needs. This leads to demanding
the unattainable from an agent and in turn creates
a pressure on the agent to fabricate.
One participant in the seminar, who has worked
as a reports officer with over 1000 agent sources
of information, said he had seldom seen a reports
officer have any power in the authentication pro-
cess vis-a-vis the operation officer. Yet the
reports officer who knows the operational back-
ground to the case should be in the best position
to evaluate the agent's information both as to
accuracy and value.' Aside from the obvious
measuring of value, regular checks on agent
information should be a normal part of the
reports officer work. A formal system to
accomplish the latter was recommended for LA
Division at one time but has not been: implemented.
If the information provided by the agent is neither
accurate, nor needed, the operation should be
halted. Tests of the accuracy and value of the
information must be met with every agent source,
no matter how many other steps in authentication
have or have not been accomplished.
Approved For Release 2002/1'1/04: CIA-RDP80-00630A000200010001-7
 CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
--the analyst: He can assist in the evaluation
of the information to assure that it meets the
two essential criteria mentioned above. But.
most analysts need a considerably better under-
standing than they now possess of how to be
more usefully critical and challenging of the
information provided in clandestine service
reporting. The seminar participants disagreed
on how much information the Operations Directorate
should provide the analyst in soliciting his coop-
eration in the evaluation of material. This led
to a criticism of the source descriptions affixed
to reports and to some comment that the Operations
Directorate should educate the analyst on the use
of source descriptions and what they really mean.
Several participants suggested that a separate
seminar in this series would be worthwhile on
the "theory and use" of source descriptions..
25X1A
-8-
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
25X1A
--the psychological advisory officers: In many
operations one or more of the variety of assess-
ments which can be made with the assistance of
the Behavioral Activities Branch are useful. in
the authentication process. But there was dis-
agreement on the importance of probing agent
motivation by psychological assessments as
part of the authentication process. One view
was that motivation was usually too obscure to
be useful in authentication. Others pointed
out that it could help in establishing the
extent of the bias in the reporting product
of the agent. Changes in motivation, which
can have a profound effect on the entire
operation, can be reflected in a continuing
series of psychological assessments.
The Importance of Definitions
It was clear to a number of the participants that some
of our difficultie-s in authentication are traceable to a lack
of uniform understanding of the meaning of elements in the
Approved For Relea fil~Wr.fIA-RDP80-00630A000200010001-7
 CONFIDENTIAL
Approved For-Release 2002/11/04: CIA-RDP80-00630 000200010001-7
process. Most of the seminar participants would agree with
the following treatment of key terms:
Authentication: The process, to be useful and correct,
is never completed and there is really no such-thing as an
authenticated source. The authentication of a source is a
continuous process engaged in during the entire period of
the agent's employment.
Agent/Collaborator: An agent is a reportiIg source who
is controlled in the classic sense following a formal recruit-
ment and who is in a formalized contractual basis with the
Agency with some remuneration. Authentication of such agents
should fully engage all the applicable steps in the process.
On the other hand, collaborators, such as some agents of
influence, high officials who have limited ideological grounds
to cooperate as a friend, etc., may constitute a growing
proportion of our intelligence sources.. They are not so
tightly "controlled" as the agent and may not have formalized
contractual agreements with the Agency. As a result, they
may not be getting as much authentication processing as
their intelligence product should require.
Tests/Assessments: Many operational tests are just
that, and are passed or failed by the agent. On the other
hand, the word "test" tends to be used when what is really
meant is an assessment--which is not a black/white process.
Assessments of all kinds can be useful even if at first
glance they tend to produce a negative image 'f`or-- the--agent.
-10-
Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
CONFIDENTIAL
 Approved For'Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
25X1A
Criteria and Guidelines: While there was a general
reaction against establishing any formal system of rigid
criteria on authentication which would be required and
reported to Headquarters, a number of seminar participants
saw the utility of developing a systematized set of guide-
lines (the individual steps of which were not compulsory,
but were comprehensive) as a management tool to focus
attention on the process and on the steps that could be
used, and to enable training in their application to be
systematized. This would be an encouragement to good
tradecraft.
-11-
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
25X1A
Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
 CONFIDENTIAL
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CENTER FOR THE STUDY OF INTELLIGENCE
12 July 1976
Factors Related to the Authentication of'Agent
Sources of Information
Do We Need Criteria?
There is considerable disagreement among case officers
as to whether criteria of general world-wide applicability
can or should be developed. Some have expressed grave
reservations about any attempt to create "dogma," fearing
it would be mechanically applied and blind us to important
individual differences in cases. One strong demonstration
that general criteria might work is provided in an article
in "Studies in Intelligence"' in which there is a striking
parallelism shown between the complaints and needs of two
such culturally different agents as a German and a Korean.
On the other hand, there are strong arguments against
2
universal criteria in the thrust of three other articles,
each about recruiting agents of different nationalities: a
25X1A
Approved For Releas?OOM1 Q1fIQLIA-RDP80-00630A00 200010001-7 25X1
 25X1A
CONFIDENTIAL ~.~
Approved IRelease 2002/11/04: CIA-RDP80-00630A000200010001-7
Soviet, a Chinese, and a Libyan. These articles suggest
that the problems typical of dealing with these three
nationalities are so different as to make similar authentica-
ttion.criteria impractical.
.,The most detailed work that has been done on the subject
of authentication of agents was produced in the context of
the Agency's mission in Vietnam and was tailored mainly to
that unique situation. This literature has found its way
into background reading for the operational training mission
of the Office of Training
although it is (probably correctly) not a formal part of the
required reading or instruction.- Despite the care with which
this work was undertaken and the direct tailoring to the
Vietnamese situation, experiences in Vietnam do not suggest
that the effort was very successful in avoiding the recruit-
ment and use of agents later found to be fabricators. The
core concepts of this effort were converted into a guideline
for the process of authenticating agents for all of the East
Asia Division in 1973 and it has been applicable there since.
What experiences has EA Division had in applying the criteria
in Stations other than Vietnam? Have certain of the criteria
;proved more valid than others and have some criteria proved
inapplicable in otherrFar East countries?
Normal Processes of Authentication:
The scattered literature available on the subject tends
to focus on some generally applicable steps in the process
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
-2-
 CONFIDENTIAL
Approved Fe Release 2002/11/04: CIA-RDP80-006'30A000200010001-7
of authenticating an agent source. The first group of such
steps is research: the identification of and development of
biographics on the individual, his documentation, the estab-
lishing of his access, the determination of what useful
information he can provide, the inter-agency and local source
checks, the Provisional Operational Approval (POA), the
Personal Record Questionnaire (PRQ) part I and part II.
Are any of these dispensable in the process of authentication?
It has been said that the PRQs have never been revised,
despite the Agency's penchant for updating all its forms.
The PRQ does not ask what steps in authentication have been
completed. Does this document need changes, and should it
contain a check list of some of the generally desirable steps
in the authentication process?
All case officers seem to concur that there is no
substitute for the case officer's thorough knowledge of the
subject matter upon which the source is going to report, and
of the agent and his personality in the fullest detail.---.Some-
case officers argue strongly that all that is needed is full,
detailed and perceptive knowledge of the agent as a human
being, and that no set criteria , for authentication are useful.
Is this belief widespread in the Agency and is it a well
founded view?
A third aspect of the process of authentication of an
agent is usually the evaluation of the information the agent
-3-
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
CONFIDENTIAL
 -) CONFIDENTIAL ti
Approved Fdr Release 2002/11/04: CIA-RDP80-00630A000200010001-7
produces. The evaluation can take many forms from checking
against known information, checking against what other sources
report, comparing with photography or evidence available in
COMINT, to submitting the information for the judgment of
analysts in Headquarters. Which of these or others are
particularly useful? How well have the judgments of analysts
in Headquarters served the purpose?
25X1A
Psychological Approaches:
On the scale of popularity with case officers, the series
of psychological tests used in authenticating agents appears
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
-4-
 } CONFIDENTIAL
Approved FbrRelease 2002/11/04: CIA-RDP80-006304000200010001-7
Such tests as those sponsored
25X1A by
the Behavioral Activities Branch, the
graphological analysis approach, and finally the analysis
of the content of the PRQ part III, are all part of the
psychological group of tests sometimes used in authentica-
tion. There are case officers who find the resulting analysis
of such tests entirely too inconclusive to be of much use.
Still others believe they are totally useless and only provide
the case officer with other ways of stating what he already does
or should know about his agent or potential agent. Are the
.results of these psychological analyses too hedged in their
language and would more unequivocal language be really useful?
There are a number of complaints and criticisms about
our general operational approach to authenticating agents
which may be worth airing in search of better methodologies.
The most common of these criticisms is that there has been
an alleged dangerous overemphasis upon recruitment in recent
years. There is disagreement as to w this is dangerous,
however. Some say that it encourages carelessness with the
proper steps of authentication (or worse, their falsification);
that contact reports are never written with firmly expressed
doubts about the authenticity of a potential recruit. Has
our training and supervisory approach instilled in the case
officer the injudiciousness of this step when a healthy
skepticism of agents is almost universally recognized as
desirable? What is causing this? Another reason cited for
the alleged danger of overemphasis upon recruitment is that
Approved For Release 2002/11/04: CIA-RDP80.-00630A000200010001-7
-5-
25X1A
 Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
it has simply failed to produce the quality of agent desirable.
The charge is that our quality agents have not, in fact, been
recruited by "gung ho"" case officers with large numbers of
recruitments to their credit. Still another explanation of
the alleged danger is that it stresses the value of the
"pitcher" recruiter) over the value of the "extractor"
(agent handler), placing the importance of the latter too
low and thus undermining the intelligence product.
Critical Points.
Some have noted that there are several critical points
involved in any vigorous authentication process. One is the
danger of probing the agent's motivation,to the point of
creating agent hostility. A second is pressing the agent
to produce to the point that he is frightened to admit he
cannot get some piece of information which is a prelude to
fabrication. There are also those case officers who believe
we have taught "arid continue to teach entirely too much to our
agents, thus creating throughout the world a cadre of well-
trained agents who know all too well how to make fabrication
believable. Is this a generally shared view? Still. another
danger is seen in too much quantifying of everything, in
terms of numbers of agents recruited, numbers of reports
turned in, numerical value of the reports, grade averages
of agents, projects, stations, brancfies and divisions, all
of which, it is charged, leaves hollow-and empty the real
-6-
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
 CONFIDENTIAL ,
Approved ForRelease 2002/11/04: CIA-RDP80-00630A000200010001-7
25X1A
value of the information collected. Is this a legitimate
criticism? Finally, a danger of particular consequence for
those engaged in hard-target recruitment is the frequent
problem of shortness of time available before a target
returns to his homeland. This can lead to the elimination
of testing and authentication steps that would otherwise be
taken.
Fabrication:
Fabrications are an endemic problem,-With-which the
Agency will have to deal so long as there is a market for
intelligence information. Little change in thinking about
this problem has taken place since the classic piece done on
the subject in the 1950's We have
apparently found from experience that some agents start out
as fabricators, but most of them seem to lose their access
or run out of information at some stage and then turn to
fabrication. Does this, really hold true, and if soy does
it imply that we are pushing the agent too hard to get what
he can't get, and/or training him so well he can invent what
we want? Is fabrication always due to the delinquency of a
case officer? Most fabrication appears to be uncovered by
25X1A 3"Fabrications, Paper Mills and Imposters," Jul- 1970
version is in OSL-2047A),
wrote the basic piece in the first half of the 1950's.
Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
CONFIDENTIAL
 } CONFIDENTIAL ",-~1
Approved Fo Release 2002/11/04: CIA-RDP80-0063DA000200010001-7
case officers in their operational dealings with the agent,
rather than by intelligence analysts pouring over the reported
information. Is this generally the case,,and__if so, why are _s_o
few fabrications tripped up on the basis of the information
itself? Is the Agency failing to provide careful enough
scrutiny of the contents of its intelligence reports for
fabrication, and if so, where should this responsibility lie
with the reports officer or the DDI analyst?
Control and Motivation:
The issue of control and motivation figures largely in
any consideration of authentication of an agent and may be
worth discussion at the seminar. Most case officers believe
a monetary motivation is best for control purposes because
this type of agent motivation is usually the easiest to
handle and exploit. The least popular control seems to be
the use of coercion; few agents controlled by coercion have
ever been successfully run for long. Between these extremes
fall other possible elements of control and motivation. The
ideological motivation is very common, but may be becoming
increasingly difficult to handle in an age when U.S. foreign
policy changes (detente) make re-recruitment repeatedly
necessary, and not always successful. Personal security,
religion, shared values with an international basis,
4
Anti-atomic warfare, disarmament, anti-biological/chemical
warfare, ecology, polution, and economic interdependence are
concepts that would fit into this category.
Approved For Release 2002/1/04: CIA-RDP80-0063OA000200010001-7
 CONFIDENTIAL
Approved F?~` Release 2002/11/04: CIA-RDP80-00630A000200010001-7
adventurism, and finally, "any motivation that may exist"
have all been mentioned. Is control and motivation undergoing
fundamental changes as the U.S. becomes a less popular moral
symbol with conflicts with the underdeveloped world in which
many of our operations are now run? Are adventurism and other
motivations likely to increase in importance for our agents,
and if so what does this mean for control?
Source Descriptions:
There is no very clearly defined point at which the
appropriate officers decide to circulate the intelligence
information provided by a source while describing that source
as untested, or "reliability not established." There is also
no agreement on when the step should be taken to describe the
source as fairly reliable, and from there to graduate him to
more confident assessments. In fact, there is an interesting
division of opinion as to whether the information or the
source is the most important thing being authenticated. One
former division reports chief believed that if the source
was trustworthy and discriminatory about his own sub-sources
of information and careful about khat he passed on to us,
then the information--regardless of what it was--should be
authenticated as reliable because it came through a reliable
channel of acquisition. Others believe that some agents may
well be incapable of discrimination about the information
that comes into their possession, and are mere couriers or
Approved For Release 2002/11704: CIA-RDP80-00630A000200010001-7
 CONFIDENTIAL
Approved Fo Release 2002/11/04: CIA-RDP80-0063DA000200010001-7
conveyors of a volume of information from which we must choose
for our best intelligence. There are obvious gradients between
these extremes. Are there useful guides to points in the
process of agent authentication when certain steps in dissemina-
tion of intelligence from an agent should be taken? Are such
guides generally applicable to all operations around the world?
There are probably other topics of significance in the
authentication of agents which have not been mentioned in
this paper, and a discussion of these will be welcomed at the
seminar.
-10-
Approved For Release 2002/11/04: CIA-RDP80-00630A000200010001-7
 25X1 Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7
Approved For Release 2002/11/04: CIA-RDP80-0063OA000200010001-7