28 February 1999. Thanks to Anonymous.
Source: Hardcopy CIRA Newsletter, Volume XXIII, Number 4, Winter
1998/1999. Published by Central Intelligence Retirees' Association, Box 1150,
Ft. Myer, Virginia 22211.
Jump to encryption remarks.
Address at the CIRA Luncheon
5 October 1998
CIRA President Maurice Sovern introduced guest speaker John Millis, Staff
Director of the House Permanent Select Committee on Intelligence. Millis
agreed to speak in place of HPSCI Chairman Porter Goss (R, Florida), who
was detained by a family emergency. Millis is a former operations officer
who spent about 12 years in the Agency, primarily in the NE Division. After
serving on the staff of the Deputy Director of NSA, he joined HPSCI in 1993,
becoming Staff Director in 1997. Noting that Millis had served "over the
ten-year mark," Sovern said in introduction, "I think I have persuaded him
to join CIRA... so his lunch today is not free."
JOHN MILLIS' SPEECH
Well, I guess the good news is that since Congressman Goss is not here, Mo
has said he is giving you each personally a 20% rebate on today's cost.
First of all, Congressman Goss thought that he would be able to be here today
but this morning he went with his wife for a checkup. It has been a difficult
time for him. He wanted me to say thank you to all of you because he has
heard from so many of you, his old friends, and he appreciates the kind thoughts
and wishes that have come his way. The good news is that his wife is doing
much better; she's going to require about four or five weeks' of convalescence,
and she has a new pacemaker. So, he's very pleased with that news but he
was very unhappy that he couldn't be here with you today.
This was a talk that he was looking forward to. This was to be a sort of
old-home week for him, many of you being his peers and colleagues from his
thirteen years in the DO, which he regards as being some of the best years
of his professional life. And he had a personal message which he wanted to
deliver to you on the topic of building a constituency for intelligence.
He thinks of you as being a core organization and you individually as being
the people that can help do this.
I'm not going to give that speech. I hope -- and he hopes -- that you will
invite him back, and when he does appear before you he will want to talk
about this proposal. I will also not try to tell you anything about the
intelligence process itself. I'm very sorry the Congressman is not here because
I will confess to being intimidated talking to you. As Maurice mentioned,
I was in the DO from '81 until '93, and you're the stars that formed the
firmament when I was a young officer. You're the giants, and I know that
there is nothing I can tell you about intelligence that you don't know better
than I do. So, what can I talk to you about? I thought that what I might
do is give you a survey of opinions: what we see from Capitol Hill; a state
of the Intelligence Community from that perspective. Now, a couple of things
first. When I've talked in this area before, I've noticed a tendency on my
part to concentrate on the negatives. That's characteristic of the Hill approach;
because if we're to do our jobs well, we need to know what's not working
well in the Intelligence Community, and believe it or not, to try and help.
To do that, we have to be able to make assessments of what works and what
I also hope that my comments aren't going to feed a tendency in some former
intelligence officers who for lament the "Good Old Days." The veil of retirement
is permeable to bad news but not much good news seems to escape out into
the broader community. If I think back to my days inside, there was an awful
lot of griping and grousing; in fact, that was (I know from my experience),
at least basically what we did best when we weren't actually working. But
we had the context of knowing as well the good work that we were doing. When
you are no longer doing the good work all you're hearing is just the noise,
the negative news. So, I hope I'm not going to feed that tendency even more
and I hope you'll keep that in mind as I speak. The fact of the matter is
despite all the negative things I'm going to say, there is some absolutely
phenomenally good work going on in the Intelligence Community. It costs a
lot of money; the American taxpayers, I'm happy to say, are getting what
they pay for, and more. I don't think you can say that about most of the
Let me start with the first piece of truly good news. It's one word: George.
George Tenet from my point of view is exactly what the doctor ordered for
CIA. You all know one of the most important things, if not THE most important
thing that a manager must do, is to pick good people to help implement his
policies and to advise him, and I think George has done a very good job of
that. If you look at the group of people that he has put together, I think
that it is a very strong team.
OVERVIEW OF THE COMMUNITY - FIVE NOTES
I want to talk about the whole Intelligence Community, because if you are
just going to talk about the CIA exclusively, you're missing the big picture.
And the big picture is what is important right now. There are five particular
areas I want to go into. These are themes that we have used on the Committee
in trying to make decisions about allocation of resources. We've taken some
rather significant actions in the Intelligence Authorization Bill that just
went to Conference. This involves, in the broadest brush of terms, taking
a great deal of money out of overhead collection systems and moving it into
other areas: to NSA, principally to modernize some of its new collection
capabilities; some of it into the CIA too, again to put a little flesh back
on the bare bones of both the analytic and the operational capabilities;
also, to rebuild analytic capabilities across the rest of the Intelligence
Community. We did that by taking some money from the NRO, but also by adding
money over and above the President's request. It's a small percentage, but
it amounts to millions of dollars that we hope will go to good purposes.
But I want to talk to you about some of the themes behind what we do with
the budget, and behind what we do in all our activities in trying to oversee
and assist the Intelligence Community.
First is to try to rectify what we believe is a continuing and historical
imbalance between investment in collection and investment in what is sometimes
called "downstream" activity: processing, exploitation and most importantly,
at the end, analysis. The second is to try to move away from the sort of
ad hoc and bureaucratically motivated investment strategies for technical
collections systems. This is a complicated point involving, one; the crisis
affecting SIGINT systems; two; a revolution taking place in imagery, caused
by capabilities in commercial imagery that have been growing exponentially;
and three; issues involving the NRO. My third major point is the need for
more espionage, to turn back to some old-fashioned spying and some new-fashioned
types of spying that basically reside in the Clandestine Service. The fourth
point is the need to rebuild covert action capabilities, and I'm thinking
here in terms of the paramilitary and also media covert action capabilities.
Finally, and this is a large a philosophical point but it's one that dominates
many of the decisions being made in the Intelligence Community: There is
a need to rebuild a strategic -- or what we sometimes call the national
capabilities -- to end what has been an absolute and total fixation on near-term,
tactical intelligence. So those are five themes I want to talk about.
COLLECTION AND DOWNSTREAM EXPLOITATION
First of all, the imbalance between collection and downstream activities:
Someone compared the Intelligence Community to a library on which we have
expended billions of dollars -- the world's best library; more volumes, more
information than any other library in the world -- and yet the reading room
is limited to a few seats. That's the Intelligence Community. The actual
cost of all-source analysis is 1% of the intelligence budget, and yet we
scrimp and we save and cut here and cut there. The idea was that we can't
afford to have DIA and the DI both doing the same sort of analysis; that
we can only afford to have one. That's ridiculous. An analyst costs less
than $200,000 a year, including salary, retirement benefits and the cost
of putting the PC's and such on the desk. We can afford many more analysts
than we have. Instead, we spend more money on one satellite in one year than
we dc on all the analytic capabilities combined. We believe that that's something
that needs to be changed.
One of the programs -- in, fact, THE program that has occupied us most for
the last six months -- is the Future Imagery Architecture (FIA). This is
a dollar system that is to give us state-of-the-art imagery for the next
twenty years. Multibillion dollars. . . This program was put together and
presented to us, and only afterwards did anybody realize that there was not
one dollar budgeted for tasking this system, for processing the data from
it, for exploiting it or for disseminating it. That struck us as strange.
Right now, well less than half of the pictures taken by our satellites ever
get looked at by human eyes or by any sort of mechanized device or computerized
device detecting change. Under this new system we will increase anywhere
between five and fifteen times the number of pictures that will be taken
from the sky. We have fewer imagery analysts than we had five years ago;
we have many fewer than we had ten years ago. And there is at present no
plan to build up an imagery analytic capability to deal with this new collection.
It's a mindset. People get fixated on hardware. We're trying to change that
INVESTMENTS IN TECHNOLOGY AND COLLECTION
My second point is what I call an ad hoc and bureaucratic approach to investments
and technology types of collection. I'm talking here of signals intelligence
and imagery intelligence. It strikes us as ad hoc and bureaucratic because
there is no discernible overreaching rationale behind it. It's ad hoc in
the sense that we buy it a la carte, system by system, each presented
individually and without relationship one to the other, without management
by the Intelligence Community. If you're hearing something that you've heard
for the last thirty years, it's because the issue is still the same. There
is still no management of the Intelligence Community. The intelligence agencies
are each managed, but there is no one in a position to make the tradeoffs
within the Intelligence Community that will make a coherent, efficient
organization that will function as a whole. So, we end up doing it on Capitol
Hill. And I've got to tell you, if you are depending on Capitol Hill to do
something as important as this, you're in trouble.
But we're often the first people who ever look at these things, the first
to say: well okay, if they're doing this in SIGINT and this in IMINT, is
there redundancy? Is there overlap? Are we better off spending a little more
money in a clandestine type SIGINT activity rather than investing a huge
amount of money to collect this type of imagery? There needs to be some authority
at the top of the Intelligence Community. This is supposed to be the task
of DCIs, and we continue to try and work with the DCI, to encourage him to
step up to this challenge. As you know, it is very difficult to exercise
authority over the National Foreign Intelligence Program and all its agencies,
because 90% of them are funded and owned and operated by the Department of
PROBLEMS AHEAD IN TECHNICAL COLLECTION
Let's talk about the problems with technical collection.
Signals intelligence is in a crisis. We have been living in the glory days
of SIGINT over the last fifty years, since World War II. SIGINT has been
and continues to be the INT of choice of the policy maker and the military
commander. They spend about four or five times as much on it as they do on
clandestine collection, and the fact of the matter is, it's there quickly
when needed. It's always there. Or it has always been there.
In the past, technology has been the friend of NSA, but in the last four
or five years technology has moved from being the friend to being the enemy
of SIGINT. You know Moore's law says that computation technologies change
every eighteen months; in telecommunications, the pace is more rapid than
that. There are 2 couple of problems in the SIGINT collection crisis. It
gets classified very quickly, but let me suggest to you a couple of ideas
to think about.
The media of telecommunications is no longer
SIGINT-friendly. It used to be. When you were doing RF signals, anybody within
range of that RF signal could receive it just as clearly as the intended
recipient. Well, we moved from that to microwaves, and people figured out
a great way to harness that as well. Well, now we're moving to media that
are very difficult to get to. Compounding this is encryption, which is about
to take off no matter what we do. This is a huge policy issue that I know
some of you have followed fairly closely. No matter what we do, encryption
is here and it's going to grow very rapidly. That is bad news for SIGINT,
so it is going to take a huge amount of money invested in new technologies
to get access and to be able to break out the information that we still need
to get from SIGINT. That money is not being invested at present. NSA continues
to do what they can do, but they need to be thinking about five years from
now and trying to invest in those new technologies.
We're spending huge amounts of money and this leads me to my third sub-point:
the agency that eats the NFIP, the National Reconnaissance Office. We spent
incredible amounts of money on overhead collection and it threatens to overwhelm
the intelligence budget. We have already spent more than twice as much money
on the NRO than on any other intelligence agency. You do both imagery and
SIGINT from satellites, but it doesn't make a lot of sense doing SIGINT from
there any more. Excepting ELINT, you shouldn't be spending one dollar more
than we do to try and intercept communications, regular voice and data-type
communications from space, but we do make that investment. This is something
that we think that we have to move away from.
Further, a revolution is coming in imagery. ADM Dantone, the first Director
of the National Imagery and Mapping Agency (NIMA), had a very good illustration
of that revolution. He said, "Picture two large TV screens behind me. On
one is an incredibly detailed picture from space showing one meter or less
resolution of a desert. You can see tank tracks; you can see tanks; you can
see ammo depots; you can see where communications lines have been laid down;
you can see communications towers; you can see where headquarters have been
set up; you can see what the direction of movement is. It's everything that
you needed to know back in 1991-92, when we were going up against Saddam
Hussein. The other screen is blank. That's what Saddam Hussein had.
We obviously had a huge advantage. Now picture two screens again. The one
over my left shoulder has a little better imagery, a little better resolution.
This is what we're going to have five years from now. And over my right shoulder
-- Saddam's screen five years from now -- is the imagery quality we had before.
There will be some leveling of the field. In the next year three US companies
will be launching satellites with close to one-meter resolution, and one
will be better than the one-meter resolution. This product will be for sale.
In the foreign arena, four countries in two or three years will be at that
same point, and they will be selling their imagery commercially to whoever
has the money to buy it. That's a big change.
The other thing we need to be wondering is why are we spending huge amounts
of money from the intelligence budget to collect pictures from overhead when
our capabilities now are well known? It's hardly an intelligence discipline
any more. It's being funded by the National Foreign Intelligence Program,
but it is not involved in collecting secrets; it's for certain surveillance
purposes. Our gut feeling is if the military needs to keep an eye constantly
on some target for strictly military purposes, we should not call this a
national intelligence system.
MORE OLD-FASHIONED ESPIONAGE
This gets to the third point: the need for more espionage. There has been
a sea change; I don't know how many of you have noticed it. Two or three
years ago, everybody was constantly beating up on the Clandestine Service
at the Agency. Now, I'm quite amazed at the change. We get lots of calls
daily from journalists, and they hardly ever call up anymore wanting just
to do a story that gratuitously beats up on the CIA's Directorate of Operations.
They're out there trying to do stories about the value. The Ames case took
a terrible toll for three or four years, poisoning the perception of the
CIA's Directorate of Operations. It seems we're getting beyond that. There
seems to be close to a consensus of opinion that we need to rebuild these
clandestine capabilities. The question is how do you go about doing it? Again,
we're very pleased by what Jack Downing has done in the DO in trying to fix
the problems and the damage which was done over the last several years. We
believe that it's going to take more resources, but I must say that we are
not inclined to throw money at the DO. We have to rebuild slowly; we have
to give graduated larger amounts of resources to the DO. We need to make
sure that it is being used as intended, not for the miscellaneous things
that they shouldn't be doing anyway. We are going to watch this closely.
REBUILDING COVERT-ACTION ASSETS
The fourth thing is the rebuilding of covert-action capabilities. Again,
without getting classified, there's not that much I can say; just a couple
of observations. We believe that we have to keep a very strong paramilitary
capability on the shelf for the Directorate of Operations. It's always hard
to anticipate what the next crisis is going to be, but we can be sure that
it will be there. And the paramilitary capability the DO has is uniquely
well suited for many of the types of situations that we're seeing around
the world. The media side was never popular among most case officers. "Where
are the metrics? How do I know that what I'm doing is of any value?" Well,
the fact is the Agency got out of the business totally. And believe it or
not, there are some situations in the last year or two where it would not
have hurt the US to try and swing world opinion our way. We see that we are
losing world opinion against Saddam Hussein. There are countries that are
being taken over by narco-traffickers. There are countries that are harboring
terrorists. It would be nice if we could do something other than deliver
a demarche or have USIS out there with the media guidance. That covert capability
is totally gone. We believe again this is something that the CIA did very
well, and has just as much application today as it did in the past.
Finally, there's a need to end the fixation we have on tactical intelligence.
Since DESERT SHIELD/DESERT STORM, we have abandoned the strategic mission
in large part to meet the pressing requirements the military has made for
tactical intelligence. We all remember Gen. Schwarzkopf's criticisms of
intelligence support, but again our memories are selective because he also
repeatedly said in a positive sense that no commander of any army had ever
had the intelligence support that he had had. There has been a tremendous
pull from the Department of Defense for more and more and more support. The
CIA stepped up to the requirement and has done outstanding job of providing
that support. ] this is an era of declining resources and people and so it
has been done at the cost of the national mission. Some people used to come
to us and brag that the CIA is the 911 of the Government. Well, if you're
dialing 911, intelligence has already lost the opportunity of doing what
it should be doing, which is telling you in advance that you are going have
a problem and helping you avoid it before it gets to the crisis.
The idea of being able to give a wonderful paper on the problem that broke
out in Kosovo yesterday is, from my point of view, not a particularly valuable
exercise, particularly for the CIA. We have to get back in the business of
having deep long-term penetration of intelligence problems, of having a strategic
look into the future.
So, these are some of the themes that we's pursued as we've made about three
or four hundred changes in the Intelligence Budget submission that came to
us this year.
CIRA Newsletter (ISSN 1096-1836) is published quarterly by CIRA at 15605
Yellowhorn Ct., Rockville, MD 20853.
To Contact the Retirement Division:
For local calls from the Washington, DC area: 703/613-8080
All calls from elsewhere in the United States: 1/800/648-5693
Address: Chief, Retirement Division, P.O. Box 9704, Rosslyn Station,
Arlington, VA 22219