Case 6:
Robert Hanssen at the FBI
Another more
familiar example of insider risk is the case of Robert Hanssen, the
FBI agent who was arrested in February 2001 for spying for the
Soviets. According to the unclassified review of his case published
by the Justice Department’s Inspector General in 2003 (DOJ, 2003),
Hanssen’s espionage career spanned three time periods: 1979-1981,
1985-1991, and 1999-2001. Several public reviews of Hanssen’s
personal and professional life have referred to the emotional
stressors that accompanied his spying during this period (for
example, see Shannon and Blackman, 2002) and the Inspector General’s
report also notes his increasing “recklessness” and
“self-destructiveness during this period (DOJ 2003: p. 14). As an
additional test of the software, we utilized Hanssen’s published
communications with his Soviet handlers (Shannon and Blackman, 2002)
over the last two periods of his spying to see if the reported
levels of anxiety, depression and general emotional vulnerability
reported could be detected across this period in these
communications. This involved applying WarmTouch to eight notes
written by Hanssen between October 1, 1985 and November 15, 2000.
The three figures below display the
WarmTouch results for several indicators of Hanssen’s emotional
state. The measures used for anger were his use of negatives (e.g.
no, not, never, etc.) and the word “me.” The measures for
emotional vulnerability—use of feelings, “I,” direct references to
others and adverbial intensifiers (terms that exaggerate adverbs
such as “so” in the phrase “I miss you so much”) were
designed to portray the direct, intense, communication of emotion to
another person. An individual’s use of terms that qualify or
discount statements (e.g., I “think” I’ll go or I “sort of”
want to go) and retract statements (like “but,” “although”) -- known
as retractors and qualifiers—were used to measure anxiety (Weintraub,
1981, 1986, 1989). As the three figures indicate, there were marked
increases in all of these measures of emotional stress over the time
period for which there was data available encompassing the last two
periods of Hanssen’s espionage.
Changes in Detected Levels of Anger,
Emotional Vulnerability and Anxiety, 1985-2000



As more and
more employee communications move on-line, so must means to detect
disgruntlement, psychological distress, and deception. As the case
of Hanssen illustrated, initial background investigations are not
useful to detect employees that become disgruntled while on the job.
In addition, disgruntled employees have done considerable damage to
organizations and national security before (and after) regular
security (even polygraph) updates. The ability to detect changes in
employee communications indicative of increased risk of violations,
narrow a field of suspects, or locate individuals with specific
characteristics or vulnerabilities from their on-line communications
could be a valuable assessment tool.
© Copyright 2009 by
Eric D. Shaw, Ph.D.
|