Date: Sat, 27 Jun 92 12:39:30 EST Errors-To: Comp-privacy Error Handler From: Computer Privacy Digest Moderator To: Comp-privacy@PICA.ARMY.MIL Subject: Computer Privacy Digest V1#056 Computer Privacy Digest Sat, 27 Jun 92 Volume 1 : Issue: 056 Today's Topics: Moderator: Dennis G. Rears Chronicle Crypto Article Re: Can Merlins be used as bugs? Re: privacy dilemma Re: SSNs and Social Insurance Numbers Re: SSNs and Social Insurance Numbers Re: FBI Digital Telephony Proposal Re: Social Security Numbers and Social Insurance Numbers The Computer Privacy Digest is a forum for discussion on the effect of technology on privacy. The digest is moderated and gatewayed into the USENET newsgroup comp.society.privacy (Moderated). Submissions should be sent to comp-privacy@pica.army.mil and administrative requests to comp-privacy-request@pica.army.mil. Back issues are available via anonymous ftp on ftp.pica.army.mil [129.139.160.200]. ---------------------------------------------------------------------- Date: Wed, 24 Jun 92 18:02:18 CDT From: Joe Abernathy Subject: Chronicle Crypto Article This cryptography article appeared Sunday, June 21. It is being forwarded to Risks as a way of giving back something to the many thoughtful participants here who helped give shape to the questions and the article. In a companion submission, I include the scanned text of the NSA's 13-page response to my interview request, which appears to be the most substantial response they've provided to date. I would like to invite feedback and discussion on the article and the NSA document. Please send comments to edtjda@chron.com Promising technology alarms government / Use of super-secret codes would block legal phone taps in FBI's crime work By JOE ABERNATHY Copyright 1992, Houston Chronicle Government police and spy agencies are trying to thwart new technology that allows conversations the feds can't tap. A form of cryptography _ the science of writing and deciphering codes _ this technology holds the promise of guaranteeing true privacy for transactions and communica tions. But an array of federal agencies is seeking to either outlaw or severely restrict its use, pointing out the potency of truly secret communications as a criminal tool. "Cryptography offers or appears to offer something that is unprecedented,'' said Whitfield Diffie, who with a Stanford University colleague devised public key cryptography,'' an easily used cryptography that is at the center of the fight. "It looks as though an individual might be able to protect information in such a way that the concerted efforts of society are not going to be able to get at it. "No safe you can procure has that property; the strongest safes won't stand an hour against oxygen lances. But cryptography may be different. I kind of understand why the police don't like it.'' The National Security Agency, whose mission is to conduct espionage against foreign governments and diplo mats, sets policy for the government on matters regarding cryptography. But the FBI is taking the most visible role. It is backing legislation that would address police fears by simply outlawing any use of secure cryptography in electronic communications. The ban would apply to cellular phones, computer networks, and the newer standard telephone equipment _ already in place in parts of Houston's phone system and expected to gain wider use nationwide. "Law enforcement needs to keep up with technology,'' said Steve Markardt, a spokesman for the FBI in Washington. "Basically what we're trying to do is just keep the status quo. We're not asking for anything more intrusive than we already have.'' He said the FBI uses electronic eavesdropping only on complex investigations involving counterterrorism, foreign intelligence, organized crime, and drugs. "In many of those,'' he said, we would not be able to succeed without the ability to lawfully intercept.'' The State and Commerce departments are limiting cryptography's spread through the use of export reviews, although many of these reviews actually are conducted by the NSA. The National Institute of Standards and Technol ogy, meanwhile, is attempting to impose a government cryptographic standard that critics charge is flawed, al though the NSA defends the standard as adequate for its intended, limited use. "It's clear that the government is unilaterally trying to implement a policy that it's developed,'' said Jim Bidzos, president of RSA Data Security, which holds a key cryptog raphy patent. "Whose policy is it, and whose interest does it serve? Don't we have a right to know what policy they're pursuing?'' Bidzos and a growing industry action group charge that the policy is crippling American business at a critical moment. The White House, Commerce Department, and NIST refused to comment. The NSA, however, agreed to answer questions posed in writing by the Houston Chronicle. Its purpose in granting the rare, if limited, access, a spokesman said, was "to give a true reflection'' of the policy being implemented by the agency. "Our feeling is that cryptography is like nitroglycerin: Use it sparingly then put it back under trusted care,'' the spokesman said. Companies ranging from telephone service providers to computer manufacturers and bankers are poised to intro duce new services and products including cryptography. Users of electronic mail and computer networks can expect to see cryptography-based privacy enhancements later this year. The technology could allow electronic voting, electronic cash transactions, and a range of geographically separated _ but secure _ business and social interactions. Not since the days before the telephone could the individual claim such a level of privacy. But law enforcement and intelligence interests fear a world in which it would be impossible to execute a wiretap or conduct espionage. "Secure cryptography widely available outside the United States clearly has an impact on national security,'' said the NSA in its 13-page response to the Chronicle. "Secure cryptography within the United States may impact law enforcement interests.'' Although Congress is now evaluating the dispute, a call by a congressional advisory panel for an open public policy debate has not yet been heeded, or even acknowledged, by the administration. The FBI nearly won the fight before anyone knew that war had been declared. Its proposal to outlaw electronic cryptography was slipped into another bill as an amend ment and nearly became law by default last year before civil liberties watchdogs exposed the move. "It's kind of scary really, the FBI proposal being consid ered as an amendment by just a few people in the Commerce Committee without really understanding the basis for it,'' said a congressional source, who requested anonymity. "For them, I'm sure it seemed innocuous, but what it represented was a fairly profound public policy position giving the government rights to basically spy on anybody and prevent people from stopping privacy infringe ments.'' This year, the FBI proposal is back in bolder, stand-alone legislation that has created a battle line with law enforce ment on one side and the technology industry and privacy advocates on the other. "It says right on its face that they want a remote government monitoring facility'' through which agents in Virginia, for instance, could just flip a switch to tap a conversation in Houston, said Dave Banisar of the Washing ton office of Computer Professionals for Social Responsibil ity. Though the bill would not change existing legal restraints on phone-tapping, it would significantly decrease the practi cal difficulty of tapping phones _ an ominous development to those who fear official assaults on personal and corporate privacy. And the proposed ban would defuse emerging technical protection against those assaults. CPSR, the point group for many issues addressing the way computers affect peoples' lives, is helping lend focus to a cryptographic counterinsurgency that has slowly grown in recent months to include such heavyweights as AT&T, DEC, GTE, IBM, Lotus, Microsoft, Southwestern Bell, and other computer and communications companies. The proposed law would ban the use of secure cryptogra phy on any message handled by a computerized communica tions network. It would further force service providers to build access points into their equipment through which the FBI _ and conceivably, any police officer at any level _ could eavesdrop on any conversation without ever leaving the comfort of headquarters. "It's an open-ended and very broad set of provisions that says the FBI can demand that standards be set that industry has to follow to ensure that (the FBI) gets access,'' said a congressional source. "Those are all code words for if they can't break in, they're going to make (cryptography) illegal. "This is one of the biggest domestic policy issues facing the country. If you make the wrong decisions, it's going to have a profound effect on privacy and security.'' The matter is being considered by the House Judiciary Committee, chaired by Rep. Jack Brooks, D-Texas, who is writing a revision to the Computer Security Act of 1987, the government's first pass at secure computing. The recent hearings on the matter produced a notable irony, when FBI Director William Sessions was forced to justify his stance against cryptography after giving opening remarks in which he called for stepped-up action to combat a rising tide of industrial espionage. Secure cryptography was designed to address such concerns. The emergence of the international marketplace is shaping much of the debate on cryptography. American firms say they can't compete under current policy, and that in fact, overseas firms are allowed to sell technology in America that American firms cannot export. "We have decided to do all further cryptographic develop ment overseas,'' said Fred B. Cohen, a noted computer scientist. "This is because if we do it here, it's against the law to export it, but if we do it there, we can still import it and sell it here. What this seems to say is that they can have it, but I can't sell it to them _ or in other words _ they get the money from our research.'' A spokeswoman for the the Software Publishers Associa tion said that such export controls will cost $3-$5 billion in direct revenue if left in place over the next five years. She noted the Commerce Department estimate that each $1 billion in direct revenue supports 20,000 jobs. The NSA denied any role in limiting the power of cryptographic schemes used by the domestic public, and said it approves 90 percent of cryptographic products referred to NSA by the Department of State for export licenses. The Commerce Department conducts its own reviews. But the agency conceded that its export approval figures refer only to products that use cryptology to authenticate a communication _ the electronic form of a signed business document _ rather than to provide privacy. The NSA, a Defense Department agency created by order of President Harry Truman to intercept and decode foreign communications, employs an army of 40,000 code-breakers. All of its work is done in secret, and it seldom responds to questions about its activities, so a large reserve of distrust exists in the technology community. NSA funding is drawn from the so-called "black budget,'' which the Defense Budget Project, a watchdog group, estimates at $16.3 billion for 1993. While the agency has always focused primarily on foreign espionage, its massive eavesdropping operation often pulls in innocent Americans, according to James Bamford, author of "The Puzzle Palace," a book focusing on the NSA's activities. Significant invasions of privacy occurred in the 1960s and 1970s, Bamford said. Much more recently, several computer network managers have acknowledged privately to the Chronicle that NSA has been given access to data transmitted on their networks _ without the knowledge of network users who may view the communications as private electronic mail. Electronic cryptology could block such interceptions of material circulating on regional networks or on Internet _ the massive international computer link. While proponents of the new technology concede the need for effective law enforcement, some question whether the espionage needs of the post-Cold War world justify the government's push to limit these electronic safeguards on privacy. "The real challenge is to get the people who can show harm to our national security by freeing up this technology to speak up and tell us what this harm is,'' said John Gillmore, one of the founders of Sun Microsystems. "When the privacy of millions of people who have cellular telephones, when the integrity of our computer networks and our PCs against viruses are up for grabs here, I think the battleground is going to be counting up the harm and in the public policy debate trying to strike a balance.'' But Vinton Cerf, one of the leading figures of the Internet community, urged that those criticizing national policy maintain perspective. "I want to ask you all to think a little bit before you totally damn parts of the United States government,'' he said. "Before you decide that some of the policies that in fact go against our grain and our natural desire for openness, before you decide those are completely wrong and unacceptable, I hope you'll give a little thought to the people who go out there and defend us in secret and do so at great risk.'' ------------------------------ From: jharuni@micrognosis.co.uk (Jonathan Haruni) Subject: Re: Can Merlins be used as bugs? Summary: Probably. Organization: Micrognosis International, London Date: Fri, 26 Jun 1992 10:15:20 GMT Daniel P. B. Smith (dpbsmith@world.std.com) writes: >Our office AT&T Merlin systems offer a built-in speakerphone mode > ... [and lots of other features besides -ed] ... >Potentially, it seems as if the system could be >used to eavesdrop on offices. The phone installer assured me that >this was impossible (translation: if there IS a way, they don't tell >him how). But I wonder if there are really engineered protections, >or whether eavesdropping is simply not a feature supported by the >"standard" software. Could a hacker reprogram it somehow? ... I worked once in an office that had an old phone system (not a PABX, but one where each extension had a button for each CO line), and every extension was a handsfree phone. Each extension could be dialled as an intercom, and the dialled extension would ring once (unless the ringer had broken :-() and then its microphone was switched on AUTOMATICALLY. In that office, nobody minded this intrusion. It was a very friendly environment with very little privacy to begin with. People were usually away from their desks, but within voice range. So when the receptionist had to pass through a call, she just dialled the extension, shouted "Oi, Joe!" and Joe shouted "WHAT?!?" without having to drop everything. These phones did have one redeeming feature, though, which was an on/off switch hardwired to the microphone. From your description of your phone system, it sounds like you have the dangerous feature (a microphone that can be switched on remotely without warning) and you don't have the safety feature (a switch). Is that true ? When you press the loudspeaker button, does it seem like a software action turns on the speaker and mic, or is it a click-on, click-off switch that is obviously local to the extension ? If it does seem like a software action, is there a separate mic control switch ? The answer to you question: If the mic is turned on by software and there is no safety switch, then OF COURSE it could be hacked to be used for spying on you. The installation engineer didn't write the software, and won't know about such things, as you pointed out. If you are in an atmosphere where you are really worried about being spied upon, and you don't care too much about using handsfree, perhaps you could arrange to shorten the life of your microphone. (Soldered joints aren't what they used to be, after all.....darn things keep breaking!!!!) But then, if people are really trying to spy on you, this will probably just pique their interest. Consider counterespionage techniques instead. Right, I'm getting carried away.... Jon. ------------------------------ From: John Artz Subject: Re: privacy dilemma Reply-To: jartz@mitre.org Organization: The MITRE Corporation, McLean, Va Date: Fri, 26 Jun 1992 13:55:57 GMT In article , samsung!ulowell!willow.ulowell.edu!welchb@uunet.uu.net writes: |> Here is a similar dilemma. We received a receipt for a property |> tax bill on a postcard (because it was a cheap, computer-generated way |> for the town). I agree that the tax assessment for all property in the |> town is publicly available, and should be. Yet, it seemed to be an |> offense against the American idea of politeness and privacy to think that |> someone could simply read my taxes off a postcard. I felt I was being |> singled out for special mistreatment, whereas I would not feel so if |> they obtained my payment from a list of all payments. |> -- |> Brendan Welch, UMass/Lowell, W1LPG, welchb@woods.ulowell.edu That is my point exactly. Nobody wants to be singled out, but if such information is readily available on everyone it becomes much less of an issue. John M. Artz, Ph.D. jartz@mitre.org +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ A crisis is just the end of an illusion. -- Gerald Weinberg +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ------------------------------ From: Mike Johnston Subject: Re: SSNs and Social Insurance Numbers Organization: Lehman Brothers Date: Fri, 26 Jun 1992 16:48:08 GMT Apparently-To: uunet!comp-society-privacy In article flint@gistdev.gist.com (Flint Pellett) writes: > If anyone really knows about SSN's, I'd love to know what plans > exist for them in the future. They only have 9 digits, and since > we have 250,000,000 people (or more-- I haven't kept track) > currently alive in this country, that indicates that most likely > 20 to 25% of the available numbers are in use by persons currently > living. I would guess that within the next 100 years that we'll > run out of 9 digit numbers that haven't already been used: do they > plan on re-using the numbers of deceased people, (a big potential > problem, I would think, since estates often live on a long time > after the person), or are they going to go to 10 digits and break > computer programs all over the place? I'm hardly an expert, but I would think that with the U.S. population expansion at a relatively low rate compared to other countries (3% I've heard quoted) it'll take a long time before the other 750 million numbers are used. If you really want to talk about software all over the place breaking, ponder that in many places a 5 digit julian scheme is used ( YYDDD ) so that dates can be relatively simply added and subtracted. What happens on on Jan 1, 2000? Yup, all of a sudden the pattern wraps around from 99365 to 00000. The results are pretty predictable. All I'll say is let's hope the SSA doesn't use this to calculate benefit elegibility.... MJ -- Michael R. Johnston mjohnsto@jenny.shearson.com Lehman Brothers (212) 464-3061 "I was a reporter, and this worried me a great deal and I could not understand how the devil I had gotten myself into such a fix." - Hesse ------------------------------ From: bear@tigger.cs.Colorado.EDU (Bear Giles) Subject: Re: SSNs and Social Insurance Numbers Organization: National Oceanic & Atmospheric Adminstration / Boulder Labs Date: Sat, 27 Jun 1992 03:07:08 GMT In article flint@gistdev.gist.com (Flint Pellett) writes: > >If anyone really knows about SSN's, I'd love to know what plans >exist for them in the future. They only have 9 digits, and since >we have 250,000,000 people (or more-- I haven't kept track) >currently alive in this country, that indicates that most likely >20 to 25% of the available numbers are in use by persons currently >living. I would guess that within the next 100 years that we'll >run out of 9 digit numbers that haven't already been used: do they >plan on re-using the numbers of deceased people, (a big potential >problem, I would think, since estates often live on a long time >after the person), or are they going to go to 10 digits and break >computer programs all over the place? If they add one digit, they'll certainly add a second (for a checksum since SSNs are used so widely now). But you make two interesting assumptions: first, that the U.S. Government will care about the fact that other organizations use the SSN as an identifier (SSN records could probably be freed within a few years of person's death); and second that the new number will only be used in the U.S. With a North American Free Trade union you could make a good point for a NAFTU-SSN... if not a global SSN (since it is likely people will change countries much more in the future). Bear Giles bear@fsl.noaa.gov ------------------------------ From: "Glenn R. Stone" Subject: Re: FBI Digital Telephony Proposal Date: 26 Jun 92 17:51:19 GMT Reply-To: glenns@eas.gatech.edu Organization: The Group W Bench In comp.society.privacy banisar@washofc.cpsr.org (Dave Banisar) writes: >The following is the latest version of the FBI Digital Telephony Proposal, >introduced in May 1992. This version removes the previous language that >authorized the FCC to set standards and now places it solely in the hands >of the Attorney General. Question: Does the Attorney General's office come under FEMA (the Federal Emergency Management As.., err, Agency)? If it does, we have REAL cause for concern.... q.v. some recent posts concerning FEMA in misc.headlines that explain how the President can basically arbitrarily declare a state of emergency and take over a whole host of things, including PTT services. If this stuff is already in place when (IF? c'mon, who are you kidding?) said emergency happens... well, use your imagination and scare yourself. (BTW, I have independent corroboration on the aforementioned articles in misc.headlines, from someone who used to have a clearance.... ) Question #2: Is someone running a lobby against this measure? If not, would someone like to? I don't have the time to run it, but I'll certainly put my John Hancock on any petition opposing it.... [Moderator's Note: Does anybody know of an organized effort against this? Anyone willing to start one? I would presume makers of the equipment are against it. _Dennis] -- Glenn R. Stone (glenns@eas.gatech.edu) I'm not in the book, ya'know, and I'm ding dang dong glad of it! -- Gopher ------------------------------ From: Hans Mulder Subject: Re: Social Security Numbers and Social Insurance Numbers Organization: University of Nijmegen, The Netherlands Date: Fri, 26 Jun 1992 23:52:54 GMT In lance@unix386.Convergent.COM (Lance Norskog) writes: >I'm pretty sure the last digit is not a check digit. Depends on whether you mean an SS# (American) or an SI# (Canadian). Both are 9 digits. The last digit of an SI# is the check digit. SS#s don't have a check digit. -- Hans Mulder hansm@cs.kun.nl ------------------------------ End of Computer Privacy Digest V1 #056 ******************************