FAS | Government Secrecy | January 2000 News ||| Index | Search | Join FAS

DoD News Briefing

Thursday, January 13, 2000 - 1:32 p.m.
Presenter: Mr. Kenneth H. Bacon, ASD PA
MR. BACON: Good afternoon. Welcome to our briefing.


Q Ken, can you comment on the Chicago Tribune story of a vast loss -- of a possible vast loss of information because of the Y2K glitch in the spy satellites, and in particular, the timing and how long you were without these systems?

MR. BACON: Yeah, I'd be glad to. The story was wrong. (Laughter.) Deputy Secretary Hamre spoke about this at considerable length the Tuesday after New Year's, and I'll be glad to repeat much of what he said. I think this is -- I don't mean to dismiss this. This is a very important issue. It's become, I think, a test of credibility between the Pentagon and the press, and I think it's very unfortunate. And frankly, I think the press is off base on this issue. Let me tell you why.

We have a very robust and overlapping intelligence capability, comprised of a number of different systems, that is designed to give us the ability to monitor events around the world 24 hours a day. These systems are, to a certain extent, redundant, and they're redundant in order to give us various ways of monitoring events around the world, but also to give us overlapping or redundant capabilities in case there is a temporary glitch in one of the systems.

So without getting into details, there are a variety of systems collecting, observing, providing information in a number of different ways.

Dr. Hamre, I think, went into considerable detail in discussing this and, in that respect, broke new ground, because we don't generally talk about our most sensitive intelligence collection systems. We made a commitment to you to reveal how our military systems were working during the Y2K turnover between December 31st and January 1st. In the context of that, we did report that we had a temporary problem with a collection system, with getting information from a collection system. We do not as a matter of course reveal problems with collection systems. We don't do that because we don't want to give away information about our capabilities or temporary lack of capabilities at any given time. I think everybody would understand that when you're dealing with intelligence collection systems of the most sensitive nature, the less said about them the better. We made an exception during the Y2K turnover because this was something that arose during the turnover.

At no time were we blinded. This has been a canard that's been thrown around in the press from day one. At no time were our intelligence collection systems blinded. That is because we have redundant systems designed precisely to deal with a variety of situations.

Two, as Dr. Hamre said right after the turnover, we did suffer a temporary interruption in information processing by one -- from one system. It was repaired quickly, in a matter of hours, although not repaired 100 percent. It took a while to get back to 100 percent capability, but we did in a relatively short period of time. During the period when we were moving from a temporary fix to the front line, normal system, we were able to, through a process of education and repetition, become so good with the temporary fix that we were restored to close to 90 percent of our total capability by the time we got the regular system back on line.

The amount of time this system was down not particularly remarkable. There are various weather and other events that can affect our systems, and this was well within the type of temporary interruption that we experience on a fairly regular basis.

As I say, you must remember we have multiple and redundant systems precisely so that we will never be blinded if there is a temporary glitch in one system.


Q Would you explain, then, when there's a temporary glitch in that system, so at absolutely no time it -- explain "blinded," I guess, is what I want you to do. You could see what you needed to see, high-priority things, even during that temporary system shutdown?

MR. BACON: Well, let me explain it in the simplest ways. In the way that most Americans would think of our intelligence systems -- that is, the ability to monitor attacks or potential attacks, or preparations for attacks against the United States -- we never lost that capability. And I think that without getting into details, that is the central fact. We always had the same type of real-time -- we always had an impressive degree of real-time monitoring capability, despite this temporary glitch.

Q But you may have lost one type of capability during that period -- optical images or radar images?

MR. BACON: Well, as Dr. Hamre said and I will repeat, we had a temporary interruption of product from a significant system, and that temporary interruption had an insignificant impact on our ability to monitor events around the world.

Q Could I ask a -- can I follow up?

Q Ken, the story, though, said two days versus a temporary -- as a matter of hours, though. I mean, there was a greater time here that the story implied here was detailed.

MR. BACON: Right.

Q I mean, you can't just blow it off as a wrong --

MR. BACON: Well, I would say the story was one part fact and two parts fiction. And the length of time listed as the interruption for this system was fictional; it was wrong.

Q Could you say --

MR. BACON: Dr. Hamre has been very clear. If you go back and read what he said on January 4th, he said: One, we had a temporary interruption. The interruption was quickly repaired.

It was repaired at about 50 percent of normal capability. Between the time it was repaired -- and it was repaired on December 31st -- between the time we put a manual fix into place and the time we got the automated system fully up and running again, we had raised our ability to manipulate the manual system to about 90 percent of normal capability. So over a period to two days, we did, I think, a Herculean job of moving up the capability of this manual fix.

Q Ken, do you mean 50 percent capability for that system; meanwhile, another system covered what you lacked?

MR. BACON: Yes. We have multiple systems.

Q But a 50 percent capability for that system; meanwhile, another system is covering what you were lacking?

MR. BACON: For this one particular system, 50 percent of capability for one of the several systems --

Q Meanwhile, you were covered by another system in what you were lacking --

MR. BACON: I am saying that we have multiple systems that all provide different types of collection. They are redundant in that you could describe the types of collection as somewhat the same in that they are intersecting circles. And what we lost was one -- if you took all our systems and put them together, they provide a total of intelligence. We lost a little corner of part of our total intelligence take for several hours. That's what happened.

Now, do we wish we hadn't lost it? Of course. But as I said, there are weather and other events that sometimes cause temporary glitches in the systems. And we did not experience any greater loss during this Y2K-related incident that we do from time to time in the normal course of operations.

Q Ken, speaking of other events, there is a piece today in Inside the Pentagon, that says that there were other technical glitches earlier in the week, and there was also a piece in the Aviation Week that talked about 1970s era computers that have been frequently breaking down in this system. So you seem to be sort of defending what happened over the weekend by saying, well, this happens frequently and this wasn't out of the norm for the kind of breakdowns we have all the time.

MR. BACON: No. No. I'd like to go back to the point I made earlier, that we don't generally talk in much detail, or any detail, about intelligence systems. I think that the reasons for that are very obvious. In this particular case, we made an exception because it was Y2K related. I think Americans would appreciate that when we're spending billions and billions of dollars on highly sophisticated, sensitive systems, that we want to say as little about these systems and their capabilities as possible so that people can't then figure out how to counter the systems in certain ways.

In this particular case, the system we're talking about, the system that was affected temporarily by a glitch on December 31st, is a highly reliable system. Every system we have, whether it's a tank, a submarine or an intelligence system, has a level of reliability. In other words, we set a standard that it is supposed to be operational and reliability X percentage of the time. This system routinely, routinely performs better than the level of reliability we have set, and the level of reliability we have set is extremely high. Now, it routinely beats the very demanding level of reliability we have set for it. But remember, this is a system that is designed to operate 24 hours a day, 365 days a year. So if something happens for two hours or three hours, it's an extremely small percentage of the time that it's supposed to be operating, or that it is operating in the course of the year.


Q Ken, when you were forced to go to the manual back-up system, was your through-put problem more an issue of amount of coverage of the Earth's surface that you wanted to look at, or was it a time -- time lag issue, more so?

MR. BACON: Neither.


Q Is it correct that the problem was caused, sort of ironically, by the Y2K fix? Is that what caused this?

MR. BACON: Well, the analogy here would be somebody on a medical procedure that had to continue all the time -- somebody in an artificial lung, for instance. A polio victim.

You would like to make a fix in the lung, but you can't stop the lung to make the fix, to test a new system. This was a system that would have required us stopping it in order to test the fix from end to end. A decision was made not to stop this particular part of the system at any time. Therefore, they never tested the patch or the fix from beginning to end. They tested it in sections, and it turned out that it was a mistake, because the sections didn't fit together, the sections of the fix. They tested each section of the fix, but they never tested them all together, and it turned out that was the mistake, that one section of the fix didn't operate compatibly with another section of the fix.

They figured this out relatively quickly and they were able to put together a manual fix and that manual fix, as I said, was up and running within a short period of time, and it was running at a fairly high level of reliability and it allowed this one system, one of several important intelligence systems, to operate -- not 100 percent, but increasingly close to 100 percent -- over the next two days.

Q Ken?


Q It seems kind of ironic to me that you guys wouldn't plan an outage of this thing to test it to make sure it would work and then suffered an outage that was unplanned later. Has there been any discussion that, heck, if we have a situation like this in the future, maybe we should instead plan when we're going to go out instead of just being surprised by it at a time when the Russian president is stepping down and potential terrorist --

MR. BACON: We are, in the Pentagon, by nature aggressive and almost obsessive planners, but we have not yet started planning for Y3K. (Laughter.) But when we do, I'm sure the people following us will take this -- will go back and check this issue.

Q Do you agree it would have been a better approach to plan an outage and plan an end-to-end test and have an outage when you can plan to have those assets covered by other assets, as opposed to just having it happen when you weren't planning on it?

MR. BACON: I think it's -- I can't answer that question. All I can tell you, I'd like to just go back to the fundamental point here. John Hamre said this glitch had insignificant results. He was right; it had insignificant results.

Q Is that a function of luck in the world situation, or something else?

MR. BACON: No, it is not a function of luck; it is a function of having highly reliable, redundant intelligence-gathering systems that give us an ability to monitor events around the world in near real time a variety of different ways.

Q Different ways, but are there redundant systems that do exactly what these satellites do -- take pictures from the sky?

MR. BACON: Well, first of all, we never talked about what this system is, publicly, and I don't plan to do it today. Second, I can just assure you and the American public again that we have highly redundant systems designed so that if there is a temporary glitch in one of them, we are not blinded or incapacitated in any way.


Q Is it correct --

MR. BACON: In any significant way. I mean, obviously, you'd always like to have everything operating 100 percent all the time. Life is hardly like that. Sometimes tape recorders and cameras fail. They're pretty low tech compared to some of the things we're talking about. But we believe that the system we've set up gives us a high degree of reliability that we can monitor events around the world to give us the types of warning we need to put our forces on alert. And that type of warning was never compromised by this event.


Q Is it correct that the faulty intelligence computer processing which -- was processing information from more than one kind of intelligence gathering system? In other words, there are a variety of kinds. Was it more than one kind that was being processed by the computer with the problem?

MR. BACON: The answer is yes and no. And I won't go any further.


Q May I change the subject?

Q One more on this?

Q Two more?

MR. BACON: Sure. Let's finish with this.

Q When Dr. Hamre said that some data was irretrievably lost in this, was that data on New Year's Eve from the time when it really went down, or was it from sometime over the whole weekend?

MR. BACON: It was most notably on New Year's Eve. There was some on other times because as we pointed out very forthrightly at the time, when the manual fix was put into place, it did not give us 100 percent of what we had before; it gave us around 50 percent to begin with, and we got it up to 90 percent. So by definition, since we weren't getting 100 percent of what we'd normally get; something was lost.

That's not the right question, though, the question you asked: Was something lost? The right question is: Was significant product lost? Did we lose product that jeopardized our ability to provide the types of information and warning we need to carry out our national defense responsibilities? And the answer to that is no, we did not jeopardize our ability to carry out our national defense responsibilities.

Q I'm sorry. Logic check. If you couldn't see the data you were getting, how do you know if it was significant or not?

MR. BACON: As I said, we have redundant systems. We have a variety of ways of getting the type of information we need. Intelligence, as you know, is -- like this briefing -- is a highly iterative process. (Laughter.) Things move incrementally over a small period of time. And much of what intelligence analysis involves is comparing long series of information, information gathered over a very long series of time. And when you have a variety of ways of gathering information, the inability to collect one particular type of information for a very small period of time frequently is not significant.

Q Ken, it's difficult to pose questions when we're not officially acknowledging the system involved. But the system involved is not primarily involved in early warning, as I understand it. The systems involved -- you can't acknowledge it from the podium, but they're not primarily early-warning, you know, nuclear-attack-in-progress systems. They're -- you know, they're image-gathering devices, the product of which is slowly and methodically looked at over time for a whole variety of reasons. So I don't -- you seem to be taking a shot at something that wasn't asserted. It was never asserted, in anything I wrote anyway, that we were exposed to nuclear attack because we were blind. The intelligence material we're talking about is long-term analytical stuff, and a picture of an --

MR. BACON: That's precisely why such a temporary interruption was insignificant. That's precisely why. And that's precisely why the obsessive interest in this is so incredibly puzzling to me. (Laughter.)

But I wanted -- the reason I mentioned early warning is I think it's important to reassure the American public, who have an expectation that our systems work that they are protected, that they, in fact, were protected. Because of our multiple redundant systems, we have a variety of ways of monitoring what's happening around the world. Some, such as the satellites that monitor missile launches are, in fact, designed to give us instant, near-instant, evidence of an attack against the United States. That system was never in jeopardy. We have other ways of monitoring what's going on around the world, and one of the reasons we have these ways is it's important for us to see if a country is assembling divisions of troops and armored equipment to move in one direction or another. That ability was not compromised by what happened in this particular situation. That was not compromised. We could still do that.

That is important to know, not because it means that -- a missile attack, of course, we have 20 or 30 minutes to respond. If a country were putting together divisions of armor to move against another country, we would have longer to respond, but we would still like to know that as soon as possible. Our ability to know that as soon as possible was not affected in a significant way by what happened to this system.

So that's why I said that American people can be confident that we, as the Department of Defense, as the nation's military, had the ability to continue the type of monitoring that we spend a lot of time and effort doing.


Q One final point, Ken. One of the things you're saying here is that it was the artificial lung and you didn't want to turn the lung off. It was so important that they didn't want to shut down the system, and yet why wouldn't they, if you have all these fabulous redundant systems?

MR. BACON: Well, I mean, it's because we just made a calculation, one, that we thought the fix was going to work. We were surprised and disappointed that it didn't work. And two, everything else being equal, we would rather have the system working. We didn't shut down other systems either. The whole idea of intelligence systems is to have stuff coming in.

But I want to point out, just repeat one more time, that there are temporary glitches in systems that occur from time to time. And this was not out of the ordinary by those standards in terms of duration, and it was not seen as a particularly significant event.

In fact, when I asked Vice Admiral Wilson about this, the head of the DIA, he described the impact as "very, very marginal." And I, in my discussions with a variety of officials involved in this, have seen nothing that would change my view of it. And I think "very, very marginal" is another way of saying what John Hamre did on January 4th, "insignificant"; "it had insignificant impact."

You know, I think one of the reasons for the fascination with this is clearly that we are not able to stand up and describe precisely what happened and precisely what impact it had. And I am sorry about that. But I just think you have to understand that there are certain things we don't want to say, certain capabilities we don't want to give away, about our systems. And it's very difficult to talk about it, even in a half-intelligible way, without giving away details that we don't want to give away.

There is a big difference between having facts, so-called facts, printed in articles or TV reports that may or may not be accurate, and having people in the government stand up and talk about them from the podium. So I am just not going to do that, and neither did John Hamre.

But I think there is another reason that the press may be fascinated by this, and that is the whole way it was revealed. And I think this is not widely understood by the public, and I am not even sure it's widely understood by all of you.

It was revealed by John Hamre. John Hamre, at a reception in his office on New Year's Eve, mentioned to a group of reporters that there was one glitch that had raised concern and we were working on it and he was not in a position to give out much detail. He made this comment approximately 45 to 60 minutes after he himself had learned about the glitch.

Of the half-dozen news organizations' representatives from which he made this comment, one news organization decided to pursue the story. We cooperated with the representative from that news organization, which ran a TV account, at approximately 1:30 in the morning on January 1st, with full cooperation -- well, I would say full cooperation, giving partial facts -- but we did cooperate on getting the story straight. And the most important part of the story, from our standpoint, was that a fix was in place, and the system was working again at some capacity. And we cooperated with that news organization. The story ran.

We got phone calls from one newspaper and two wire services and one other TV network after the story ran. We cooperated with the representatives of those news organizations.

And one other TV organization ran a story on the -- early in the morning on January 1st. Two wire services ran the story. And the newspaper ran the story two days later.

Despite this, this somehow has gained the fantasy that we were holding back information. We were not. It was Dr. Hamre who revealed that there had been a problem in the first place. He spoke about it at great length on January 1st, at a briefing he gave here at noon -- at noon -- and he has spoken about it at great length since. And I am speaking about it at great length today because I hope it'll go away. (Laughter.)

To my mind, the most fascinating aspect of this is, it's a story about how the press operates. It's a story about jealousies among members of the press. It is a story about people not acknowledging how this came out in the first place, and that we have been working hard, within the admitted limits of our ability to discuss intelligence systems, exactly what happened here. And our primary responsibility, as Dr. Hamre said, is to protect important systems that protect our security, and that's we've tried to do.

But within the context of Y2K, we did talk about one particular temporary glitch in one of several intelligence-gathering systems.


Q Change of subject?

Q I'd like to say something first, before we change the subject.

MR. BACON: Sure.

Q You did overlook one important fact about the sequence of events that night, which is that Admiral Willard, who was in charge of the Y2K monitoring, stood right where you are at 9:30 that night, after this event had happened, and made no mention of it, and assured everybody that there had been no glitches --

MR. BACON: Right. First of all, as we said, as Dr. Hamre said on January 1st, Admiral Willard did not know that the event had taken place until after his briefing was over. Within an hour of Dr. Hamre's knowledge of the briefing, he had admitted -- he had revealed it in a conversation with several reporters, and we were answering questions to anybody who called and asked about it. We did not --

Q Why did not the admiral in charge of giving out Y2K information know?

MR. BACON: Typically --

Q Was he not briefed deliberately so he couldn't tell us?

MR. BACON: No. That is a churlish and I think objectionable suggestion.

Typically, if a glitch like this were to occur in the system we're discussing, it would not be reported until the next day. In other words, if a glitch -- this glitch -- we discovered this glitch at approximately 7:00 in the evening -- 1900. They knew -- they had sort of figured out -- they knew for a fact that it was a glitch by about 7:30 -- 1930. They reported it two hours or so after that. My understanding, from talking to people engaged in the system, is that normally they would not report a temporary glitch like this until the next day when it would be reported in just a sort of the average daily after-action report.

Now, if I'm wrong on that, I'll correct it, but that's what I've been informed by members of the system. The reason this was reported, and the reason it received the amount of attention it got was precisely because we had stood-up a Y2K operation to examine how all our systems were operating. So the report came, as Dr. Hamre said. The team worked aggressively to fix it; that's what they were concentrating on. The report was made in a relatively timely manner; it wasn't made instantly. But the report didn't reach the National Military Command Center until Admiral Willard had nearly completed his briefing, and he didn't learn about it until shortly after his briefing.

Now, then the question you could ask is, okay, given that, should we have called everybody back and had another briefing at 2:00 in the morning when we knew what had happened, when we had figured out what was going to happen, because as John Hamre said, we clearly were not going to stand up and announce that one intelligence system was temporarily blinded; it would have been dumb, and we try not to do dumb things. So we weren't going to say anything about it until we knew we had in a fix in place, and indeed, that's when we did talk about it, when we had a fix in place several hours after we discovered what the problem was.

If you want to fault us on not calling everybody up at 2:00 in the morning and saying, "Oh, by the way, we're going to report to you on a problem we fixed," I will accept the blame for that. But I don't think that that was a conscious effort to cover this up.

And to deal with your question, I think John Hamre and Admiral Willard both dealt with it on January 1st and again on January 4th when they briefed on this. They made it very clear that at the time Admiral Willard briefed, he did not know about this incident.

Q Do you think Dr. Hamre made it clear on January 1st that the problem was not totally fixed yet, leaving us to find out Monday morning that it hadn't been fixed till Sunday night?

MR. BACON: Well, I've gone back and read the transcript, and I believe what he said was -- I don't have it here -- that we had a manual fix in place and we were continuing to work on the total fix. And I'll go back and review that, but of course you can review it, too. But I think that he was pretty clear that we were working on getting a permanent fix in place.

Yes, Pam?

Q Yes, one final thing. If I can sling an arrow back at you, the reason that I worked on this story and why it came up on my screen two weeks after is not because I'm churlish and jealous, but because people who were actually working the problem say it's far more serious than what the Pentagon let on, and indeed the impact was insignificant, but that was a fact of the way events unfolded that night. And there is a lot of concern that the money isn't being spent to upgrade this stuff.

So, you know, don't stand up there and tell us that we're all, you know, motivated by completely selfish things. I report. That's what I do for a living. I don't sit around --

MR. BACON: Oh, I understand that.

Q -- (inaudible) -- everyone else is doing.

MR. BACON: I think that the -- I don't think I described anything you said as churlish and objectionable. It was in response to something that -- (laughter) -- that John said, which was that -- (laughter) -- and maybe I should talk about you in the same sentence, but I don't think I was referring to anything more than what John said. And I --

Q Just let the record show that I object to the idea that we're all just jealous of --

MR. BACON: I stick by my description of his comment as churlish and objectionable, that particular comment. That's not to say everything he says fits into that.

But I think, again, you know, John Hamre, I, anybody who talks about national security, we have to try to find a balance between openness, on the one hand, and protection, on the other hand. I think we found the right balance here. This is a very complex system. It was a complex incident. It was one where people worked very hard to cure the problem quickly and did. The bottom line is that it did not have a significant impact on our intelligence-gathering capability. I've tried to put that into perspective.

I've tried to explain to you that there are times when there can be interruptions in service, and this was not, in terms of the time itself, out of the ordinary. There were things that were out of the ordinary. The fact that it happened because of a Y2K switchover. The fact that we weren't able immediately to return to our main line automated system did cause us some inconvenience. We've admitted that we did lose some product, but we don't think it had a significant impact on our ability to monitor events around the world.


FAS | Government Secrecy | January 2000 News ||| Index | Search | Join FAS