February 25th at 7 o'clock p.m. and all day Saturday and Sunday February 26 and
27th Boris Kogarlitsky will deliver a Russia update seminar. Boris Kogarlitsky
is a founding member of the Russian Party of Labor and one of the best-known
figures in the growing Russian Democratic Left. He will discuss
conditions in Russia today with an emphasis on the radicalization of the
labor movement. The Friday evening lecture will take place at the Fashion
Industry High School 225 West 24th Street between 7th and 8th avenues in
Manhattan. For information about the Saturday and Sunday seminar call 212-242-4201
212-242-4201. That's 212-242-4201.
And this is radio station WBAI in New York. Time for Off the Hook.
I cut myself while shaving. Now I can't make a cuff. It couldn't get much worse. But if they could, they would.
For Billy Bone, for the best, expect the worst. I hope that's understood. For Billy Bone!
And this is Off the Hook, the program that talks about advances in high
technology, computer systems, telephone systems, what have you. All the latest in
the decade of technological fun which is upon us. And tonight we're going to be
talking about something we've talked about before, namely the upcoming Clipper
chip or whatever it's being called this particular week. We'll be having somebody
from the National Institute of Standards and Technology taking your
questions and explaining things from a slightly different perspective. So stay
with us.
Ever since it was proposed last April, the Clipper chip has caused all kinds of
controversy. This is the answer to criminals using telephones to get away
with crime, to keep their identities secret. This is a way that people can be
monitored, perhaps. All kinds of concerns are being raised. We've addressed them
here before on this particular program. And we'll be addressing them again, I
guess, as time goes by. Tonight, however, we have on the phone with us Lynn
McNulty from the National Institute of Standards and Technology, known as NIST.
He's the Associate Director for Computer Security. And Lynn McNulty, are you there?
Yes, indeed. Okay. Perhaps we can start by getting a brief history as
to what NIST is, first of all. Then we'll get into the Clipper chip from there.
Well, NIST is the new name for an organization that's been around since
the early part of this century. We were created in about 1910-18, somewhere
around there, as the National Bureau of Standards. And actually, the
importance of standards goes as far back as the framers of the Constitution, who
basically ascribed one of the roles the federal government would play would be
to establish standards. And so in the legislation that created the Bureau of
Standards in 1908, we were given the role of doing standards. It was
primarily envisioned as a physical measurement at that time. But as the
electronic age has gone along, we've gotten into setting standards with
respect to computer systems and many other types of non-physical related
events or activities that we get involved in in today's world. I see. Now,
as far as the actual Clipper chip, how did NIST become involved in that?
Well, we became involved in it a couple of years ago, after a piece of
legislation was passed in 1987 called the Computer Security Act of 1987, that
gave NIST the responsibility for providing leadership for the security of
the federal government's unclassified systems. And that's an area where a lot
of what we do to protect the government's unclassified systems has a
lot of that direct applicability into the non-government, the private sector,
and also to state and local governments. And we were faced with addressing some
of the issues that have been out there for quite a while with respect to
civilian cryptography. And at that time, our leadership identified a
potential problem that where whatever we might do in the area of providing a very
effective civilian cryptography standards would potentially have an
adverse impact upon the law enforcement community, as well as the national
security community, in terms of inhibiting their ability to conduct
lawfully authorized wiretaps or electronic surveillance activities in
the United States. When did that start becoming a concern, that wiretaps
would somehow be impeded? I'm sorry, I missed the first part of your question. When did
that begin to become a concern? I can't give you a specific time and place. I
would say within the last three to four years. I think within that last
three to four years, certainly the discussion, the recognition of all of
the implications of the digital age and how encryption might become
pervasive in that age, in terms of not only copper wire voice
communications, but cellular computer communications, all the different ways
that people were beginning to communicate. Not only that, databases, the
protection of databases, and people started talking about data hostage
issues, where potentially somebody could encrypt an entire database and then walk
out the door and hold it ransom, and possibly expect somebody to pay them a
large sum of money in return for the single cryptographic key that would
unlock a corporate or a government agency's database. Well, in a world with
such a thing as cryptology, how do you counter something like that? Is it
even possible to counter something like that? Well, it may not be, but
certainly the objective of what our agency was, was not to publish
something as a government standard that would do great harm to our
nation's law enforcement activities. And I think that was one of the
need to try and strike a balance between the people's requirements
for cryptographic protection in communications and for database
protection, as well as the need to ensure that we didn't do a lot of
harm to the law enforcement community in this country, and the need to try and
find something that had some sort of a balancing impact to it. Now,
whether you agree that what we've done with issuing the
clipper chip achieves that balance or not, but that was the underlying
principle that guided the decision makers through the discussions
that ultimately led to the decision by the part of the president to
approve the clipper chip initiative. Now, it was first brought up
last April, I believe. Yes, on April the 16th, I think, of 1993. Now, how much
work had gone into this beforehand to come up with the clipper proposal? Well,
I'd say that there was several years' worth of work. Probably in 1989, we
requested the assistance of the National Security Agency in developing a suite of
civilian cryptographic algorithms. We started on an electronic signature
algorithm that's commonly referred to as a digital signature algorithm, and it was
always our intent to go down the road from the digital signature algorithm and
ultimately work on something that would address and provide stronger
cryptography on a link-to-link basis without, at the same time,
impeding the ability of law enforcement and national security people to
execute lawfully authorized electronic surveillance activities. And in some
cases, the process was speeded along by the announcement in September of
1992 of a commercial telephone security product that the AT&T had developed. And
in this product, they had a cryptographic algorithm that was based
upon another one of our standards, the data encryption standard, that provided
excellent security and did not allow government access to real-time
conversations when that access was properly authorized by a
court order. And so a lot of our activities became focused around trying
to develop something that would work in the AT&T telephone security device and
that would, again, get back to this balance between answering, providing
people with a need for effective, high-quality cryptographic protection
while at the same time not locking out law enforcement agencies from executing
lawfully authorized surveillance. So AT&T was more or less involved in the
development of this? Very much so, because AT&T had developed this product without a lot of
government input. Basically, the government became aware that
they were developing this product, I think, in the summer of 1992. They
announced it at a convention in Texas in September of 1992. It's an
excellent product. It provides outstanding voice quality. It provides
unique session cryptographic keyings so that every time, for example, if you and I are
using the device, every time that we establish a cryptographic session,
there's a brand new cryptographic key that's used to encrypt our voice
conversations, which makes it very difficult for anybody that is
intercepting those communications and trying to feed those into a
supercomputer to obtain the decryption key, because every time you are able to
decrypt the conversation, the next one you're on the outside looking out again.
You're on the outside trying to find the key again, and that makes it very
difficult if you have a lot of conversations to process or to monitor.
The Clipper chip was proposed to AT&T. However, before anybody could agree to it,
it had to go through a lot of political vetting at the highest levels of
government. That was somewhat complicated by a change of administrations.
However, when the Clinton administration came into office, they were
briefed on the issue. The president ultimately approved the Clipper chip initiative,
and it was publicly announced on the 16th of April. I think a day later,
on April 17th, AT&T announced that they had agreed to put the Clipper chip into their
telephone security device. That's where things have been rolling ever since last April.
Would you say that the Clinton administration, because they're a bit more
technologically savvy, were they more into this Clipper chip? I know it originated
with the Bush administration, but four more years of Bush, would this have happened?
I'm not in a position to make those kind of judgments. I haven't been in those
kind of meetings where I've been able to view the decision makers and make those
kind of judgments. But I think you probably have a good insight, because
obviously this administration is much more tuned in on electronic system and electronic
communication systems related issues. Clearly the National Information Infrastructure,
otherwise known as the Super Data Highway, is a high priority project for them.
I think they understand the implications of both the good and the negative implications
of cryptography and are at least wrestling with the problem and attempting to provide
leadership for the nation in this particular thorny issue that comes with the explosion
in electronic communications that we are about to embark upon.
Let's talk particulars. First of all, where did Clipper get its name?
I've heard the people from the National Security Agency, who in a sense took the lead
after we had requested assistance from them, that developed the chip, basically said that
one day somebody was sitting around the office and they said,
what do we call this thing? Somebody said, I've always wanted to call it the Clipper chip,
and that's where it came from, as far as I know.
No particular rhyme or reason, it's just somebody's particular whimsy about what they've always
wanted to do in terms of naming a piece of silicon after, I guess, the kind of Chesapeake
Bay ships that used to run up and down Chesapeake Bay back in the days of sailing ships.
Wow, that's kind of a surprise. I didn't actually expect it to come about in that particular way.
I was expecting some great, significant insight on that, but that's about the best I can do,
and I think that's pretty close to being accurate.
Now, as far as the chip itself, how exactly is this thing going to work, in layman's terms, as possible?
Let me walk you through it, and if I get too complicated, please interrupt,
and I'll try to make it as non-technical as possible.
Each chip, we're talking the Clipper chip here, and there's another chip that is being developed
that's called the Capstone, but they all perform the same thing.
Each chip that is manufactured and that it is programmed, in other words, the logic is read onto the chip,
each chip has a unique chip identifier that is programmed into it.
The chip identifier is called the chip unique key.
There is also an encryption algorithm that is also programmed onto the chip that is 80 bits in length,
and that is called the skipjack algorithm that is programmed onto this chip.
When you and I want to establish a unique cryptographic session with our telephone boxes,
we basically push a simple red button that is marked secure.
Between my telephone device and your telephone device, a unique cryptographic session key is developed.
This session key is in turn encrypted with a chip unique key that corresponds to the serial number of the chip.
There's another piece of data that is also programmed onto the chip that is called the law enforcement access field.
Each time that there is a cryptographic session established,
the law enforcement access field is communicated down the line.
The law enforcement access field is basically the chip unique key plus the chip ID number,
which in turn is encrypted with something called the family key.
In this case, it would be the Clipper family key.
Let's assume that you and I are engaged in some criminal activity.
The law enforcement people having a warrant to monitor my communications or yours would pick up the signal.
It would appear to them to be white noise.
They, in turn, would feed it into a black box that says,
okay, that black box would decrypt the law enforcement access field.
It says, okay, this is coming from a Clipper-equipped telephone, and here is the unique serial number for the chip.
Assuming they had a court order to start with, that would allow them to go to the two escrow agents,
which were announced two Fridays ago or so, actually on the 4th of February,
as being NIST, my organization, as well as a non-law enforcement component of the Department of Treasury.
They would provide the law enforcement people with two components, the split parts of the key that they hold.
When these were combined, that would give the law enforcement people on the scene the ability to unlock the chip-unique key,
which would allow them to gain access in real time to unlock the cryptographic session key,
which would give them real-time access to plain text voice that was being communicated between myself and yourself.
I hope I've done that in about as non-technical a way as I can.
Well, it's pretty complicated.
It's a fairly complex thing, and it takes people sitting down and reading a couple of times before they really understand what's going on.
But I've tried to communicate it in a way that I hope a good many of your listeners were able to understand it.
Well, I know that one of the main concerns is in such a scenario where law enforcement officials have these two pieces,
and they piece them together, and they're able to monitor phone calls.
What's to stop them from continuing to do that well into the future?
Do you have to pretty much get a new phone?
Again, I'm not a lawyer, but my friends in the Justice Department that I've come to know and worked with over the past year
have given me a good education on how electronic surveillance warrants work.
Basically, a warrant is issued by a judge or some legal authority,
and it's good for a certain period of time.
At that point in time, the law enforcement personnel must terminate the surveillance.
Now, in the process that we're setting up, we are going to reinforce that procedural requirement
with an electronic one where we hope that ultimately when the time duration of the period of electronic surveillance expires,
the system will automatically zero out the key that has been used to unlock my conversations and yours,
and the law enforcement people will be on the outside looking at only white noise again,
and they will not be able to, and the only way they'll be able to gain access to our conversation again
is to go back and get another court order issued by a properly constituted legal authority.
Well, that brings up the next question, then.
How is this data being transferred from one agency to the next?
Is it on a disk? Is it on a piece of paper?
Well, right now we are using secure telephone devices,
but ultimately it will be used, there will be encrypted data links that will be fed right into the decryption device,
and all that information, the two key components, which when married together obtain the CHIP unique key,
will be encrypted themselves, will be transmitted themselves in encrypted form,
and they will never be seen physically by the law enforcement people.
They'll just be fed indirectly to the decryption device,
and they will be combined mathematically to provide, to develop the CHIP unique key,
and at the time that the court order times out, then the encryption key will be zeroed out,
and the law enforcement people will be on the outside and will be unable to decipher any of the encrypted communications.
And nobody will be able to, say, monitor that phone call where they got that particular bit of data and duplicate that in the future?
No.
I see.
I mean, that's our ultimate objective for the system we are developing,
and we'll be deploying within the next, you know, as we go through Phase 1, Phase 2, Phase 3 of our deployment of this technology.
Uh-huh.
Well, I mean, it sounds like a mathematician's dream here, all this encryption and algorithms flying around like crazy here.
Are there any concerns on your part as far as potential holes in the security?
Well, we've looked at it very hard from a, you know, kind of done some reverse engineering
or look at it as if we're the weak links, how do we attack it if we had to attack it?
And I think we feel very comfortable with the design we're developing.
Now, we're not deploying that technology yet.
We're starting out in a very simple phase that has a high level of security.
We'd be using secure telephones and facsimile machines that would be encrypted with high-quality encryption.
But there is some paper that would be created, and our ultimate goal is to eliminate the paper,
to eliminate the potential for access on the part of law enforcement people to any of the key components
that would potentially allow some of the abuses that maybe you're thinking about to take place.
We want to eliminate the potential for any abuse.
We want to make this a system that the individual citizens can have a great deal of confidence in
in terms of that it conforms to the explicit letter of the law.
Now, you don't believe that there's any real potential for abuse on the local law enforcement level.
Is that what you're saying?
Well, I think in some respects this may even curb abuses at the local law enforcement level.
I mean, if there is any abuse potentially of electronic surveillance, it may be at the local law enforcement level.
And, frankly, the way the procedures have been set up for local law enforcement people
have to go through various levels within their state and within the federal government to obtain access to the keys.
And there would be monitoring, after-action auditing to make sure that people are not abusing their authorities.
And, again, if we ultimately deploy the system where a lot of the securities
and the electronic devices that we deploy out there,
and these would automatically time out after the authorization of electronic surveillance,
then, again, the potential for rogue law enforcement people to be able to continue a surveillance pass when it's authorized,
I think, is greatly diminished.
Particularly if you have a lot of encryption deployed out there,
that would really thwart anybody that is trying to operate outside of the express legal authorization to conduct wiretap.
They just simply would not be able to get the keys to unlock the cryptography that would be in these telephone devices.
Now, just for the sake of argument,
supposing there was some sort of a breach at a very high level of either your organization
or another organization that had access to a lot of this information.
Now, if such a breach were to become apparent,
would that mean the entire system would have to be restarted from scratch,
or is there some sort of provision for such a breach?
Well, I mean, obviously, as evidenced by the problem that came to light at the CIA yesterday,
there's always a potential for people in high places to go bad.
But we think, in this case, we're enforcing a split-knowledge type of scenario.
We're enforcing an area where you have to have massive levels of collusion to carry something like that out.
And we just think that, frankly, people that have a high degree of integrity,
there will be a high level of oversight built into this.
And the possibility of that is very, very remote.
And we believe that this system, if fair-minded and independent people look at it,
they'll be able to come away with a good feeling that the fundamental values of the Constitution of the United States
have been preserved in this particular application.
Do you think, though, if it was just one agency as opposed to two that was guarding this information,
would you have as much faith in it then?
No, I wouldn't.
I mean, that's one of the reasons why we've gone with the split-key type of approach,
to make sure that we are enforcing collusion, that no single individual or organization holds the keys.
That it requires two people performing illegal acts,
and hopefully that kind of collusion will be quickly identified and tracked down and will be known very quickly.
Now, I know that there's a move to export this technology as well,
to be using this particular standard. Is that correct?
Well, part of the announcement that was made on February 4th pertained to export controls.
And what the government said was that it was prepared to ease export controls on key escrow encryption technology.
Now, that somewhat begs the question whether any other government will let it into their country or not,
in a cryptographic system where it's well-known and everybody knows that the American government
would potentially have the access to the keys to unlock the encrypted voice that was being encrypted over it.
But I think the fundamental problem that I talked about at the beginning of this session is not unique to the U.S. government.
There are people all over the world that are wrestling with the impact that electronic technology is having
upon the ability of law enforcement agencies and national security agencies
to perform public safety and national sovereignty type of functions.
For example, I think two years ago in Europe they were about to deploy a cellular technology
that all of a sudden the police in Europe woke up to would complicate their mission of trying to enforce criminal laws in Europe.
And the level of cryptographic protection was greatly reduced from, say, maybe a level of 40 bits down to about 16 or 20.
Well, in the case of the U.S., we've chosen not to try and mandate weak cryptography for the commercial sector and individual citizens.
What we have proposed is a piece of very, very strong cryptography, stronger even than the one we currently use as the government standard,
the data encryption standard.
But what we have chosen to do is to couple the strong cryptography with the law enforcement access field
that is surrounded by a bodyguard of laws that start all the way from the Constitution and go into our civil and criminal codes.
So I don't think that what we're seeing in the United States is going to be unique to any of the other advanced technological countries around the world.
It's not debated as publicly as we do or have the long, prolonged discussion of it that we're going through.
And whether they happen to adopt a specific type of technology, this key escrowing technology that's been developed in the U.S.,
I think it's very premature to say, and I don't know whether any discussions have been held on an official government-to-government basis.
Well, the reason I brought it up was not so much whether or not they would buy it or accept our ideas.
I was more concerned with if they did use a technology like this, let's say they weren't a democratic government, what could they do?
What could a totalitarian regime do with this kind of technology?
Could they, say, for instance, be able to tell where people are at any particular time of the day or night
simply by turning a switch and seeing where their phone is emanating from?
I don't believe that's the case, but they certainly could do traffic analysis.
They certainly could do very extensive monitoring, provided a monitoring of calls,
which in many cases they do now with telephone systems that aren't encrypted.
I think there's a lot of stories coming out about monitoring practices that were in the former Soviet Union and other states within their orbit.
But we throw encryption on that, but at the same time you encrypt conversations,
you give, say, a totalitarian government the ability to slice through that encryption.
Yeah, that's a real danger there.
I think we're back to basically why are we bothering to encrypt,
because if the government has access to conversations in real time,
then the cryptography, no matter how strong it is, in the wrong hands is a fairly transparent, protective measure.
One particular example that I can recall is in Bucharest in Romania, where a lot of conversations were tape-recorded,
and because of the nature of tape-recording conversations,
they simply collected all these tapes but never had the chance to actually listen to them.
Now, with this particular technology, does it not make it a lot simpler to store,
a lot simpler to access, to perhaps scan for certain patterns or traffic analysis, as you said?
I don't think that... I mean, I think you're back to the same problem again.
You have to have the technology to process the conversations.
Encryption doesn't... the ability to gain access to encrypted voice conversations
certainly doesn't solve that problem for you.
You have to put the technology in place that would either take those tapes
or take the direct digital feed in and process it, screen it, filter it, look for keywords, things like that.
If you don't have that database, then the ability to gain direct access to encrypted calls doesn't buy you a whole lot.
Well, I mean, I was under the impression such technology does exist or is on the verge of being created.
I'm sure it does, and it depends upon the state of technology in any given country
and the willingness of a totalitarian government to make those kinds of investments,
to sacrifice some things to make investments in that kind of technology.
I think, you know, with respect to the kind of discussion we're having about the United States,
I don't know whether this conversation is that relevant or not.
Okay. Well, we're talking with Lynn McNulty, the Associate Director for Computer Security
at the National Institute of Standards and Technology
and one of the organizations behind the now-famous Clipper chip.
Now you know how it got its name.
We're taking phone calls now at 212-279-3400.
And, Lynn, is there anything you'd like to say before we bring on the calls?
Well, I'd like to say that, frankly, this is an issue that shouldn't be primarily viewed as just a technical issue.
I think it's a very fundamental public policy issue
and one that basically says a lot about what kind of a society we want to have as we proceed into the information age.
Do we want a society where our law enforcement people are put at a disadvantage?
Or do we want a society that attempts to balance individual privacy
and recognizes the fact that law enforcement and national security people also have a job to do
and shouldn't be handicapped by swinging the pendulum so far to one side of the spectrum
that they are hindered in performing their legal function?
Okay. Let's go to our first call.
This is off the hook. You're on the air.
Good evening. Yes, I was fascinated by the comment you made.
Could you speak up a little, please?
I'll change phones. Just a moment.
Okay.
I was fascinated by the comment you made concerning the ease with which the two agencies that would hold a key
could communicate that to an enforcement authority.
At any given moment, there are going to be thousands of warrants that are needed for monitoring of illegal activity.
It's not as if the local police are going to have to go to NIST or to Washington.
They're going to have to go to some local body.
There's going to have to be a network of access control or something like that,
which means it's going to be in the hands of a lot of local people, isn't it?
Well, let me take exception to one thing that you talked about right at the beginning of your question.
You said at any given time there would be thousands of wiretaps going.
I don't think the figures bear that out.
Over the last few years, we've averaged in this country about 900 criminal justice wiretaps a year,
and that's state, local, and federal.
That's 900 for the whole country,
and about another 900 that have been authorized under something called the Foreign Intelligence Surveillance Act.
We've had at most maybe 1,900 wiretaps going at any time in the whole year at all levels of government.
But, of course, every wiretap involves actually physically traipsing out to the phone line and attaching something.
It's a bit of a bother, isn't it?
Well, sure, but obviously in these days where you're talking digital bitstreams,
there are ways to direct a bitstream to a remote monitoring site.
Again, all of this assumes that we have a legally authorized warrant to perform these.
But I think that, frankly, the kind of system we're seeing and we're going to evolve into
the components that are held in trust by the two escrow agents will be basically transmitted in encrypted form
and will be entered into the decryption box in electronic form
and will not be seen in any kind of visual or paper-based by the people that are actually,
the law enforcement agents that are actually conducting the search.
We can do that.
I think that process will not involve going through chain and layer after layer of official
to where it could be potentially compromised somewhere along the way.
Do you have any other questions, caller?
Yes, just one other comment.
That is the transmission of those two parts of the standard into the decryption box.
Has presumably the idea of that signal transmission being intercepted, recorded, and retransmitted has been addressed?
Well, I think that's one of the reasons why we're looking at even encrypting the components for transmission
so that specifically the kind of scenario you pose could not happen.
Well, who holds the key to that particular encryption?
I'm sorry, what?
Who would be holding the key to that, to that data that's being transmitted?
You'd have to be able to unlock it somehow.
Well, you know, obviously with the types of communication devices that are out there now,
you can establish a unique, I mean, using public key cryptography,
I can establish a unique session key for each time I transmit that data.
There's no paper trail or no record created.
I mean, it happens and there's nothing.
And then at the end of the transmission, the key disappears.
All right, thanks for your call.
Thank you.
One question that has come up repeatedly,
and maybe you can finally put an answer to this,
is the Clipper chip simply for telephone calls or could it be all-encompassing data transmissions, email?
Well, the concept of key escrowing is intended, I think, ultimately to be all-encompassing.
But the Clipper chip and the one that is,
the key escrowing concept that is embodied in the standard that was approved by the Secretary of Commerce on February the 4th
is primarily focused upon telephone transmissions of data, fax, and voice in a telephone switch arrangement.
We have not yet gotten into the data world or not.
And that's where basically the capstone chip is intended to fill that area.
And how we handle that is in a standards contest or not, we haven't really started thinking that one through or not.
Yeah, you mentioned that before.
Right now the Clipper chip works in the, and has been incorporated into the AT&T Secure Telephone product.
And the standard was written around that device or others that may, made by other vendors that may conform to that
to provide voice security or security for transmissions, modem transmissions over telephone lines.
You mentioned the capstone chip before. Is that what it is, capstone?
Yep, capstone, right.
Any story behind that name?
I honestly don't know. I'll try and find out.
Just for the record.
I don't know where that came from at all.
Now, where does this stand, the capstone chip, and what are its capabilities supposed to be?
Well, the capabilities of the capstone chip are a multifunction chip.
It provides stream encryption or link encryption, but it can also be used to just,
it has the electronic signature component on it that could be used strictly to sign documents
without having to further encrypt them, basically for authentication integrity of electronically transmitted documents.
And you wouldn't even have to encrypt them for confidentiality purposes.
So it has a couple of functions, one of which is a signature capability.
It has an encryption function built onto it.
And it also comes with the key escrow encryption as a free good as well, in that case.
And does it have the same capability of being unlocked?
That's right.
I see.
It has the skipjack algorithm on it, the same algorithm that is in the Clipper chip.
And do you have any kind of a timetable on this?
I really don't.
We're not, that program, the capstone chip is originally,
or was initially being developed for use in the Defense Department's defense messaging system.
And I'm not clear when that system will go out for a bid to be developed
and delivered by some contractor that will be contracted for by the Department of Defense.
Okay, let's take another phone call.
Good evening.
You're on off the hook.
I have a question regarding Clipper chip.
Sure.
Am I on the air?
Yes, indeed.
Excuse me?
Yes, you are.
Go ahead.
Okay.
And that is, what's to prevent people from using other more secure means,
particularly I think the RSA, the RIVUS, SHMER?
Well, there's nothing that would prevent them.
I mean, and this is an important point that I should have mentioned earlier in the evening.
The standard that was promulgated on February the 4th is a voluntary standard.
It's voluntary for federal agencies.
It's voluntary for the individual citizenry in this country,
and it's voluntary for individuals in the private sector.
Basically what it does, though, it specifies the key escrowing technology.
It confers standards blessing upon the skipjack algorithm,
and that connotes government approval
and a certain amount of government endorsement of the strength of that algorithm.
But it is a voluntary algorithm.
There's no licensing.
There's no registration.
There's no mandate that you must use it to do anything with,
as far as the standard that has been promulgated two or three weeks ago.
If that is the case, then, and somebody develops another commercial program
that becomes very, very popular that the government does not have the key for,
will there be any reaction to that?
Well, I can't predict that.
I mean, certainly that would evoke, and assuming it became very popular,
that would evoke some level of concern on the part of our law enforcement community,
and they might potentially ask for some sort of leveling in the playing field.
But at this point in time, there is no thought on the part of the government
to make any kind of regulation or legislation with respect to cryptography.
But you don't see that as completely impossible in the future?
I think that anything that I could say is pure speculation.
I really don't want to set off any kind of headlines that are unwarranted at this point in time.
Well, I was a cryptographer, or had cryptography classwork as a graduate student in electrical engineering,
and one of the principles that we always had was that the standard and the algorithm
should always be public, and I'm sure that one of the things that's going to happen
is that as soon as the standard comes out,
there are going to be people who do try to figure out what the algorithm is.
What's going to happen if, say, someone does come out independent of official clearances
and knowledge of what was behind it, comes out with that and says,
here's the way they're doing it?
I honestly can't tell you.
In that case, we'd have to go back to the drawing boards.
Our senior managers have been assured by the people who developed this at the National Security Agency
that the potential for reverse engineering was highly unlikely
and that the security that attended the workings of this chip
and the ability to sever the law enforcement access field from the encryption algorithm
was very remote at this point in time.
But obviously, that is a potential risk that the people in government that made this decision
believe is a manageable risk and have made the decisions on that basis.
See, the other thing I sort of wonder about is there are public access encryption fields out there now,
in particular, for example, the Rivas-Shamir-Edelman,
where the actual algorithm has been published and is widely available both in the U.S. and overseas.
That's exactly true.
Let's take the case of the AT&T telephone security device.
You had a device that was user-friendly.
You had two buttons to push, and it was very easy to use.
It was intended to be sold as a piece of consumer electronics equipment
that would be sold in AT&T phone centers, Kmarts, Walmarts,
other kinds of electronic consumer outlets.
And I don't think we've seen that kind of distribution in consumer availability of high-quality,
easy-to-use cryptography up to the point where what was being talked about for the AT&T telephone security device.
See, right now, Apple, IBM, a whole bunch of people have signed on to the RSA algorithm
as what they're planning to use.
But that's going to be more in a data communications environment,
whereas we're looking at the voice environment.
And I agree with you.
I mean, I was at the RSA conference in January out in California,
and I must say it was very impressive the number of people that have signed up to put RSA in their product line.
Presumably that will be packaged in a user-friendly way.
Sure.
It's going to be, A, mean that easy to use, and B, it also means that presumably,
when you get into the whole question of export controls, for example, I mean, once it's here, it's over there.
No, you're correct.
I mean, there aren't any easy solutions to this problem because for every case you can cite,
I can cite a different one.
So there are a lot of complexities to this whole issue.
And, again, the government has gone down a road that it believes by offering up the key escrowing,
in a voluntary fashion it believes that responsible people will make decisions based upon a recognition of the need
to strike some sort of a balance on the encryption issue here.
Thanks for your call.
We're going to move on.
All right.
We have a question concerning Clipper.
Has anyone actually committed to using this chip in a commercial product, apart from AT&T?
I'm not aware of anybody at this point in time, but it doesn't mean that somebody hasn't agreed to use it,
but there hasn't been any public announcement made that I'm aware of at this point in time.
Okay.
Let's take another phone call.
Good evening.
You're on off the hook.
Yes.
Hello?
Yes.
Go ahead.
Yeah.
Hollyanna?
I must say I never have, rarely have I heard such profound naivete in my life.
What in the world makes you think that law enforcement agencies are going to behave in a strictly legally authorized manner
when we know that the history proves that quite frequently that's exactly what they don't do,
behave in a legally authorized manner.
And, you know, it's just beyond absurd to think that that's going to happen now
just because they have this new method of accessing, you know, voice communication through this Clipper chip.
And what I think we're looking at here is potentially or possibly, you know,
as things progress in the future and as this becomes more widespread and more used,
the potential to create a kind of environment or society in which, you know,
privacy is a thing of the past and where there is increased monitoring,
particularly the fact that the NSA is involved in this project makes it even more frightening
because that's the agency that has a very long history of no regard whatsoever for the Constitution
and for privacy or any laws.
I believe the NSA has always been used in foreign projects.
They're charters in foreign intelligence collection.
Did they have to get some sort of a note or something to engage in this?
I'm sorry, what?
Well, I mean, they're doing something that's not foreign here.
The NSA has a dual mission.
One is in the area of, obviously, foreign intelligence collection,
and the second mission is in the area of protecting government communications,
particularly those in the national security community.
And the Computer Security Act of 87 specifically authorizes us, NIST,
to draw upon the technical expertise at NSA in discharging our functions
that were assigned to us under the Computer Security Act.
I don't know what to say to the gentleman that called in.
He certainly reflects a point of view that is out there.
Many, many people that I have talked to have indicated
or have voiced variations of that particular theme,
and I think that's certainly a point of view that people like myself
that are out talking on behalf of this program are aware of,
and we certainly try and do our best to handle, in an honest manner,
the type of questions that may come up from people that hold that point of view.
How would you address that issue,
a worst-case scenario of law enforcement abusing people's rights?
I don't even think it's a worst-case scenario.
I think it's something that's going to be routine,
because we know that from historical facts that that's what happens.
Well, all I can tell you is that the legal safeguards that surround us
both procedurally and from a technical standpoint
have been reviewed at the highest levels of government.
Well, that doesn't reassure me.
Well, I'm sorry, but, you know,
the sense is that the Attorney General of the United States has looked at this,
and the highest elected officials in this country, the executive branch,
have also reviewed this and have also approved it.
Are you familiar with the various criminal activities
that the Attorney General of the United States have often been involved in?
Right, well, I'm sorry.
I guess you are, otherwise you sound very intelligent.
I would think you probably are familiar with that.
Oh, certainly.
I mean, I've lived in Washington for 20 years or 25 years or so,
but I think the people that are involved in this program believe that we are,
this combines adequate constitutional and legal safeguards
to prevent the kind of scenario that you are projecting from ever happening.
Thanks for the call.
Good luck.
Good luck to all of us, I guess.
Let's see if we can squeeze in one or two more phone calls.
We're talking to Lynn McNulty from NIST about the Clipper chip.
Good evening. You're on the air.
Yes, good evening.
I have some comments and some criticism, I have to say,
because I think the issue is being blown out quite out of proportion.
From the very first floater that I read on the New York Times,
it was transparent to me the intent of the industry, the computer industry,
to go all out against what they perceive as a financial loss
resulting from the lack of government contracts,
because that really is the issue.
It is not mandatory to put this chip in the computers.
Nobody has to use it in any case,
unless they want to do business with the government
or export to so-called sensitive countries
where the technology may be used for so-called terrorism and things like this.
So I question your intent,
because this program to me is a forum for ideas to come out,
and I wonder why, and if you're conscious of what you're doing,
giving all this space to this thing.
I will hang up and listen to your feedback. Thank you.
Okay.
Is that directed at me or you?
I wasn't sure, to be honest, who he was talking to.
Well, to answer if it was directed to me,
this is a forum to discuss controversial issues of the time,
and certainly the Clipper chip is one of those,
and I think we're trying to represent all viewpoints here
so that we can know exactly where this is coming from,
where it may be headed.
I certainly think that obviously maybe in a future show
that there's obviously room for somebody that is involved
in some of the public policy discussions
that represents a contrary point of view to come on
and fully elaborate upon those particular views that are out there.
Right. Well, we've had that quite a few times in the past,
and probably will in the future as well.
We've given out a lot of information concerning that,
and I think it's good to hear all sides in this issue
because you learn all kinds of interesting things,
and you can read between the lines if you feel like it
and see all kinds of other things.
Let's see if we have time for one more phone call.
Good evening. You're on.
Okay. Hold on. I'm going to lower my radio.
Good idea.
Okay. Seems like a TH1138 reporting.
A THX, that is.
Okay.
So it looks very similar to the PROMIS software.
We're all going to have a trap door to be spied on,
like that former Attorney General Ed Meese had sold
to all the countries around the world from the Iran-Contra affair.
Uh-huh.
Do you have a question?
No, that's what it just seems like to me,
and I understand Europe is not doing that.
How does this prevent anyone
from getting all the encrypting devices from Europe
and putting it in their phones
instead of what America is going to be forced to have?
Well, you know, right now I question the availability
of whether Europe is making the kind of high-quality,
relatively low-cost encryption devices
that can be readily exported from European countries
and imported into this country.
I don't think that most all of the advanced countries
around this world are going to be wrestling with the same problem
when it comes to civilian or commercial cryptography,
and I think that you're going to see this argument and debate
break out in a lot of different countries around the world.
Okay. Let's see if we can get one more phone call in.
Good evening. You're on.
Yes.
My big concern is sort of a sidetrack of the
what are people going to do if they don't want to use Clipper
or use third-party items.
I think there's an economic issue here.
I mean, if people are not going to trust the Clipper
because they know there's a trapdoor in it,
especially people in other countries
or people who deal with other countries
and don't want anybody having any kind of access
or, let's say, social action groups,
it's going to present a tremendous economic advantage
to anybody who wants to market a third-party device.
I mean, somebody from Japan, let's say.
You know, this basically seems like we're undercutting ourselves
by making this thing vulnerable.
The second issue is that at some point
during the manufacture of those chips,
the two original keys are together
and then they're distributed to these two separate agencies.
So what's to prevent some kind of corruption
at the place where the two keys are generated in the first place?
Well, let me handle the second question first
and then I'll go back to the first one.
We have developed very tight, rigid security procedures.
The escrow agents that are there at the time that the chips are programmed,
they monitor the programming of the chips,
they take immediate custody of the key components
after the chips have been programmed.
They supervise the sanitization of the programming equipment.
So, you know, we hopefully have minimized the exposure on that.
On the first question,
I think, I can't remember what that was,
but pertaining to the foreign availability,
oh no, the disadvantaging of U.S. vendors,
that point has also been brought to a lot of attention,
to the government's attention by a lot of the correspondence
that the private sector vendors have generated.
And I, you know, don't, we don't believe
that they've made an effective argument along that case,
that there will be a disadvantaging.
Who knows, it's hard to look at the future with a 2024 sight.
Okay, well we are out of time.
Lynn McNulty, I want to thank you very much
for taking the time to talk to us and our listeners here tonight.
Lynn from the Associate Director for Computer Security
at the National Institute of Standards and Technology.
This is a discussion that I'm sure will be continuing in the future
concerning the Clipper chip and its implications
and where it's going to be headed.
Again, thanks for being with us tonight.
Thank you, I've enjoyed it very much.
Okay, and this is off the hook, this has been off the hook.
We'll be back again next week, next Wednesday night from 10 to 11,
talking about some other issue of technology.
Until then, stay good and stay tuned for the news.
Good night.
Now I can't make a call
It couldn't get much worse
But if they could, they would
Bum-diddly-bum for the best, expect the worst
I hope that's understood
Bum-diddly-bum
This is Anthony Sloan for the Pitts WBAI Radio Project.
And I want to invite you to the Martinson Theater
at the Joseph Papp Public Theater this Monday, February 28th
for a live presentation of 12 Points of Light for Blackness.
This will be presented live at the Martinson Theater
at the Joseph Papp Public Theater
this Monday from 3 to 4 in the afternoon.
I want you to join Tyree Gerard, Vince Williams and others
as we present this collage of music and words
to affirm the contributions of Africans
in this wilderness of North America.
So come on down to the Public Theater live
at 425 Lafayette Street, right below Astor Place,
or you can listen to us live on WBAI 99.5 on your FM radio dial.
Or it would be best if you just come out and fill the seats
because it is free and we want you to enjoy
this fitting end to Black History Month 1994.
Good evening.
In the news tonight, the ceasefire in Bosnia is holding
and there is hope it will spread.
There's an arrest in the firebombing
of a Philadelphia community newspaper.
And in New York, the World Trade Center trial goes to the jury.
With these and other stories, I'm Jose Santiago in New York
with Verna Avery-Brown in Washington
and this is the news for Wednesday, February 23rd, 1994.
First to Washington for the Pacifica National News.
The U.S. today delivered a formal protest to Russia
over the case of a CIA official accused of spying for Moscow.
President Clinton said the spy scandal was very serious
but wouldn't warrant a change of U.S. policy toward Russia.
I do not think the facts of this case this time
undermine in any way, shape, or form the policy we have followed
for the last year toward President Yeltsin and his government
and the forces of change in Russia.
I do not believe that.
But this is a very serious case
and it has to be pursued aggressively and we will do that.
But Senator Dennis DeConcini,
chairman of the Senate Intelligence Committee today,
proposed that U.S. aid to Russia be frozen
until the U.S. can assess the damage of the espionage.
The U.S. government revealed yesterday
that it had arrested CIA veteran Aldrich Ames
and his wife on charges of spying for the former USSR and Russia.
A former U.S. intelligence officer said today
that two Soviets who were spying for the United States
had been killed because Ames passed their names on to the Kremlin.
The Russian parliament today gave President Boris Yeltsin
a slap in the face by voting to pardon
the jailed leaders of last October's rebellion.
The amnesty is supposed to take effect immediately.
Today's resolution also pardons those who staged a coup
against former Soviet leader Mikhail Gorbachev in 1991
and Russian businessmen convicted of anti-communist activities
no longer considered crimes in post-Soviet Russia.
Bosnian Muslims and Croats reportedly have signed a ceasefire agreement.
State TV in neighboring Croatia says military commanders
for the two sides signed the agreement in Croatia's capital.
It's to take effect Friday.
Previous ceasefires have collapsed,
but lately there's been heightened pressure
on the Muslims and Croats to silence their guns.
That's because a two-week truce between Muslims and Serbs
is still holding around the Bosnian capital, Sarajevo.
The Muslim-led government and the Croats
had been allies against the Serbs
before they turned to fighting each other.
President Clinton today released a revised version of his forest plan.
Emily Harris has details from Portland, Oregon.
The environmental impact statement of the ten options Clinton is offering
is the last public proposal before the administration...