If you get something from COMINT, you generally trust it. But, problems:
1. if part a says something to party b, doesn't mean it's the full truth, he may have limited
understanding or trying to trick party b
2. communications intelligence gives first rate information on third rate issues – a lot are
irrelevant to you. Rarely are they messages of great importance. Best source is the best kind of
agent, who can steal material for you – rare. COMINT has a high proportion of material
between 50-80 percentile of significance.
3. Viewing the external characteristics of communications – so communication can be designed to
be misleading, through traffic analysis
4. sigint is vulnerable to being spoofed and deceived – if you adopt traffic analysis as a long term
method of monitoring, you become vulnerable to someone manipulating their traffic
5. standard problem with agents is uncertain reliability – tell you a lot, hard to be sure of what is
accurate, but COMINT give trustworthy material, although not necessarily about things you
want to know, and certain components of it and SIGINT can be deceived.
Artillery intelligence – WWI, following the sound/source of gunfire.
These technical forms of intelligence, such as SIGINT and RADINT (radar), are highly technical,
provide massive amounts of data that need to be processed effectively. Military orgs have become
extremely competent at this processing. These forms give good intel on narrow issues – military forces
in high intensity war that are not necessarily of value outside of that. Cold war used a lot of telemetry
to track the rivals nuclear weapons.
Images acquired from overhead surveillance.
Dated back to 1915, early stages of WWI.Aircraft for reconnaissance. First few months, armies are
stunned to find radio interception and aerial reconnaissance provide amazing information. Cameras
begin to be put on aircraft – high altitude photographs. WWI, because aircraft can't go long distance,
imagery provides millions of images of the frontline. 20 million surviving photographs from
belligerents from the first world war. Have to process that amount of information – identify what's in
photos, making it useful to commanders. Can get good picture of enemies on the front.
Aircraft become longer range, imagery becomes more powerful. In 1939, figured out how to pack
larger high quality cameras on aircraft. Could get 10000 images at a time. Continues in the Cold War
with the U2. Can provide hundreds of thousands of images of targets tens of thousands of miles away
from its base. Lately has moved towards satellite imagery – resolution goes up steadily, now possible to
use googlemaps to see military class imagery in 1989, which would've been altogether possible in
Intelligence officers may use google earth for way to get imagery or location data on someone they;re
looking into. Imagery as a source has both power and weakness: if you can overfly the target you want
to observe, you can have live view of what's on the ground. Combo of SIGINT and imagery -
“something weird going on”, prompting imagery mission. Or imagery can give SIGINT hints of what
to look at. Problem is it's a huge “vacuum cleaner” - taking in so much data you can't manage it, by the
1980s. The real dilemma for imagery and SIGINT is how many people do you have to process data?
With imagery, you have enough people at any stage to look at the top priorities, but everything under
that is too much. End up going through it month by month, falling months behind. Once the surprise has happened, immediately that becomes high priority and you turn your analysts onto the right
material and they can sift through the right area of it. Dilemma with imagery and lower grade SIGINT
is processing the material you've got. Issue becomes how does one prioritize scarce resources for
examining material? Can learn through experience, but it's the distinction between the data you can
collect and that which you can assess which creates the biggest dilemmas in intelligence. Even if you
have the information and material to keep you from being caught off guard, it can still happen. Most
intelligence failures because the material was there and not a high enough priority, or your enemy turns
your system against you.Camofluage, for example.
Strategic surprise or intelligence failure i