April 1993 - cypherpunks-legacy

CLIP: Revised sample letter
by zane＠genesis.mcs.com 20 Apr '93

20 Apr '93

I finished my letter about the wiretap chip. (I still need to proof for grammar, but the content I have set.) I'm posting it here so others may see it and their efforts could be aided. Feel free to distribute this AS MUCH as possible. If you want to comment on the letter, so so publicly, because I am sending this tomorrow morning and sending to me privately will do no good-- I will have sent it by the time I read your comments. Comments *will* be helpful though for other people. I will keep the lists informed as to what response I get from my letter if people wish. -- | Sameer Parekh-zane(a)genesis.MCS.COM-PFA related mail to pfa(a)genesis.MCS.COM | | Apprentice Philosopher, Writer, Physicist, Healer, Programmer, Lover, more | | "Be God" - Me __ "Specialization is for Insects" - Robert A. Heinlein ____/ Editor: The Clinton administration on Friday unveiled their plan for establishing a standard data encryption system for voice communications. President Clinton says that he wants to bring the United States into the twenty-first century. This proposal is bringing us to 1984. I will mention first the technical reasons why the system is inadequate. The encryption algorithim is classified; only a select group of people will be allowed to examine the algorithim for flaws. The members of the cryptographic community continually and persistently emphasize that the only way to ensure security in a cryptographic system is to have as many people as possible analyze and try to break it for as long as possible. A system which has been examined by a small segment of the population should not be trusted. Release of the algorithim is crucial to verification of a good encryption method. The earlier Data Encryption Standard (DES) for data storage encryption was a very strong standard; the academic world examined it and after a number of months found weaknesses, spawning the now-standard "triple-DES" system which is more secure. From what little is known about the encryption system, it appears to be a weak system. Such a weak system lends itself to easy decryption by an unauthorized party. It would lend a false sense of security to laypersons in the field who do not realize that a key of such simplicity could be cracked easily by any talented criminal. The necessarily secure communications between a doctor and his patient could be thus breached. If the system were strong, the government would use it for internal use, but according to the AT&T release, the government will not be using the same chip which is marketed to consumers. Apart from the technical flaws in the system, there are many political problems with this big brother proposal. First, there is the assumption that the government has a right to spy on its own citizens. The proposal for this wiretap chip includes the registration of keys with two escrow agencies. This proposal is purported to allow law enforcement to keep track of "terrorists" and "drug-dealers." The first flaw in this key-escrow system is that no self-respecting criminal will use a cryptography system which can be easily tapped by law enforcement officials-- they will use strong cryptography. The only people who may end up using the wiretap encryption system will be law-abiding laypeople who don't fully understand cryptography. (Law-abiding citizens who do understand cryptography will use strong cryptography to preserve their privacy from a talented criminal.) The proposal states that in order to obtain the key of a wiretap chip user a law enforcement agency must first establish that they have a valid interest in the key. Translated out of legalese, that means that all a government agency will have to do to get access to all of the private communications, for example, between a lawyer and her client will be to fill out the necessary forms. Registering cryptographic keys with the government is similar to giving the IRS the keys to your house and filing cabinet. A criminal who wants access to the communications between a priest and confessor needs only to find a corrupt judge. The chip is being manufactured exclusively by one company. The release stated that the Attorney shall request (i.e. coerce) telecommunication product manufacturers to use this product. This aspect of the system is a government-mandated monopoly. Such monopolies result in high prices and the elimination of market forces which drive the improvement of technology. (One needs only look at the state of the Soviet Union to see how the lack of market forces affects consumer technology.) The system exposes our President's hypocrasy because of his campaign promise to protect womens' rights to privacy and that he will see a Supreme Court nominee who believes that the Bill of Rights guarantees a right to privacy. By mandating a weak cryptosystem he is reneging on his promise to provide privacy rights to the nation's citizenry. If Clinton supported a right to privacy to limit government interference with regards to abortions, he must limit government interference with regards to communication. Another element of Clinton's hypocrasy lies in his promise to reduce the budget deficit. By introducing additional responsibilities for government agencies (keeping track of the millions of keys registered in escrow) he is only using our tax dollars to invade our privacy, tax dollars which are better spent lowering the budget deficit. What I fear most from the proposal is that if the wiretap chip becomes the standard, strong cryptography will be declared illegal. If such is the case, then only criminals will have access to strong cryptography. As I have stated above-- the wiretap chip will not be used by criminals because of the obvious flaws in the crypto-system-- criminals will use strong crypto, while law-abiding citizens will have to use a system which can be easily defeated by any criminal. Strong cryptography already exists for data communications, for -free-. Strong cryptography for voice communications for -free- is only a few months away for people who own a personal computer. There is no way that making strong cryptography illegal will stop it-- it will only turn otherwise law-abiding citizens into criminals. Sincerely, Sameer Parekh 829 Paddock Lane Libertyville, IL 60048-3743 zane(a)genesis.mcs.com 708-362-9659

1 0

NPR piece on "Clipper Chip" this morning
by robichau＠lambda.msfc.nasa.gov 20 Apr '93

20 Apr '93

NPR did a (for them, anyway) well-balanced piece on the wiretap chip this morning. Per their standard, the reporter (Dan Charles) had sound bites from both sides: Whit Diffie, representing the strong-privacy-through-crypto crowd, and the (acting?) director of NIST, Raymond whose-last-name-I-forget. Highlights: Diffie compared Clipper to a real estate lockbox. The feds don't have to have the key to the house (=phone), just the key to the lockbox. If you change your Clipper key, the chip keeps a copy. NIST guy said that he strongly supports individual privacy, but law enforcement needs have to be counted, too. Diffie (rough quote): "Technology makes policy. If the gov't spends hundreds of millions of dollars on a chip which allows them to tap phones, they will do so because the technology's there." Good for NPR. A balanced piece. -Paul -- Paul Robichaux, KD4JZG | HELP STOP THE BIG BROTHER CHIP! NTI Mission Software Development Div. | RIPEM key on request.

1 0

Re: True Randoms
by jrk＠information-systems.east-anglia.ac.uk 20 Apr '93

20 Apr '93

I have seen mentioned in sci.electronics (not recently) that true random number generator chips are available off-the-shelf. Texas Instruments was mentioned as one source, but I don't remember details. -- ____ Richard Kennaway \ _/__ School of Information Systems Internet: jrk(a)sys.uea.ac.uk \X / University of East Anglia uucp: ...mcsun!ukc!uea-sys!jrk \/ Norwich NR4 7TJ, U.K.

1 0

Re: Sound bite time
by tcmay＠netcom.com 20 Apr '93

20 Apr '93

Phil Karn, referring to some parallels with gun control, writes: >I really, *REALLY* hope that this argument doesn't catch on. >Regardless of your opinions on gun control, you have to admit that >linking crypto to weapons saddles it with an enormous amount of >political baggage that we simply doesn't need. And it plays right into >the hands of those in the government who consider it as a "munition" >for export control purposes. > >I think we already have plenty of strong arguments in defense of the >right to encrypt without opening up this can of worms. It can only >turn off a lot of people who would otherwise support us. I agree. Mentioning guns raises too many confusing issues, including emotional reactions, talk of private vs. public ownership, etc. Several decades ago it might have been better received. The compelling "sound bites" revolve around these kinds of examples: - having to deposit copies of all your private letters with the authorities, "in case" they need to later read them... - not being allowed to use locks--on anything--that the government can't bypass - recording all private conversations and escrowing the tapes, just in case the government later wants to hear them - video cameras in all bedrooms to allow the police to check for illegal activities (even with proper "court orders," most people will react with shock at this suggestion) These are things that everyone can understand. And be shocked by. When they realize just how similar the "key escrow" idea is to these examples, they may get more indignant. -Tim May -- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, smashing of governments. Higher Power: 2^756839 | Public Key: MailSafe and PGP available. Waco Massacre + Big Brother Wiretap Chip = A Nazi Regime

1 0

Re: Sound bite time
by karn＠qualcomm.com 20 Apr '93

20 Apr '93

Paul E. Baclace: >As we move into the electronic frontier, the freedom to use crypto-privacy >technology is becoming the equivalent to the right to bear arms: >it is the last line of defense against a tyrannic government. The good >news is that privacy is a defensive technology, not an offensive one. >Giving up this un-enumerated right could be disasterous to future >generations. I really, *REALLY* hope that this argument doesn't catch on. Regardless of your opinions on gun control, you have to admit that linking crypto to weapons saddles it with an enormous amount of political baggage that we simply doesn't need. And it plays right into the hands of those in the government who consider it as a "munition" for export control purposes. I think we already have plenty of strong arguments in defense of the right to encrypt without opening up this can of worms. It can only turn off a lot of people who would otherwise support us. Phil

1 0

Sound bite (a bit crude)
by Douglas Barnes 20 Apr '93

20 Apr '93

Came up during brainstorm session for article: "Clipper Chip Encryption: A leaky condom for the masses?" -- Doug (gumby(a)wixer.bga.com)

1 0

Is 40 bits too little?
by Hal 20 Apr '93

20 Apr '93

Just to remind everyone, the clipper key is not 40 bits. It is 80 bits. It is not divided up into two 40-bit halves. It is the result of xor'ing two 80-bit numbers, each of which would be held separately. Knowing one of the numbers would not improve your chances of guessing the key. Hal

1 0

Re: Is 40 bits too little?
by wcs＠anchor.ho.att.com 20 Apr '93

20 Apr '93

h Yes, 40 bits would be too little, especially since you can probably tell if you've got the correct key by the form of the data, but they're actually generating your 80-bit key as the XOR of two other 80-bit numbers, and searching 2**80 still takes reasonably long. Unfortunately, the method of generating the two 80-bit numbers is disturbingly suspect; see articles in sci.crypt and alt.privacy.clipper posted by Steve Bellovin and somebody with email from Martin Hellman and Dorothy Denning. Bill Stewart

1 0

Is 40 bits too little?
by Alexander Chislenko 20 Apr '93

20 Apr '93

Yes, in just a trillion iterations you can test all possible keys. But if for testing *each* key you will have to figure out whether a text /signal you produced represents sensible speech... - it may be prohibitively expensive. Alex Chislenko.

2 1

Sound Bite
by uni＠acs.bu.edu 20 Apr '93

20 Apr '93

"Giving the government keys to your encrypted messages is like giving them keys to the front door of your house." + "and accepting the promise, we'll only use them with proper authority."

2 1