Verisign gets export approval
Forwarded from PGP-USERS list:
First PGPInc and now VeriSign? Hmmm. Is this telling us something?
"VeriSign on Monday said it received permission from the U.S. Department of Commerce to export 128-bit strong encryption software and issue digital identifications to approved organizations based on that software. " "Under the 128-bit scheme approved by the U.S. government Monday, companies will not need to place their encryption keys in escrow, or submit to U.S. government key-recovery requirements in order to use VeriSign's software, company officials said." # Thanks; Bill # Bill Stewart, +1-415-442-2215 stewarts@ix.netcom.com # You can get PGP outside the US at ftp.ox.ac.uk/pub/crypto/pgp # (If this is a mailing list or news, please Cc: me on replies. Thanks.)
At 09:57 PM 7/16/97 -0700, Bill Stewart wrote:
Forwarded from PGP-USERS list:
First PGPInc and now VeriSign? Hmmm. Is this telling us something?
"VeriSign on Monday said it received permission from the U.S. Department of Commerce to export 128-bit strong encryption software and issue digital identifications to approved organizations based on that software. "
It tells us that the US government has found yet another sucker to support their failed policy of bait and switch. VeriSign, just as AT&T and National Semiconductor have discovered in the past, will discover soon that the revenue generated by "playing ball" isn't nearly as large as promised. [How many Clipper phones and Fortezza iPower cards were sold? Total?] In fact, it the revenue may well prove to be in the negative digits. Here is the straight dope on the VeriSign/MSFT/NSCP/USG deal: If you are 1. A non US-bank (the feds decide what constitutes a bank) and promise to be nice or 2. A US corporation with a server inside the US and thereby subject to subpoena of all records then VeriSign will issue you a special cert, subject to veto by the feds, that will enable exportable Netscape and Microsoft browsers to connect to your site with 128 bit SSL. The cert is typically valid for a year, but is subject to revocation at any time by VeriSign upon the USG's request. Such revocation or refusal to issue a new cert after the first year of operation will leave the webserver operator with a server that is no longer able to encrypt communications to their customers in any meaningful way, thereby effectively shutting down Internet based operations of the company unfortunate enough to invest in such a flawed solution. In other words, the USG now permits you to use strong crypto in web based communications with your international customers if you agree to play by the USG's rules and allow the feds to install a MASTER-OFF switch in the heart of your business. What is most amusing from the government's perspective is that once the USG flips the switch, it will be VeriSign, Microsoft, and Netscape that take the heat for selling their customers such a flawed solution. --Lucky Green <shamrock@netcom.com> PGP encrypted mail preferred. DES is dead! Please join in breaking RC5-56. http://rc5.distributed.net/
This is sort of entertaining considering that Verisign still doesn't use strong encryption for themselves. (Set your browser to not use broken algorithms, and try to connect to www.verisign.com. Makes me want to send them the data for a class 3 cert.) Lucky Green wrote: | > | > "VeriSign on Monday said it received permission from | > the U.S. Department of Commerce to export 128-bit | > strong encryption software and issue digital | > identifications to approved organizations based on | > that software. " | | It tells us that the US government has found yet another sucker to support | their failed policy of bait and switch. VeriSign, just as AT&T and National | Semiconductor have discovered in the past, will discover soon that the | revenue generated by "playing ball" isn't nearly as large as promised. [How | many Clipper phones and Fortezza iPower cards were sold? Total?] In fact, | it the revenue may well prove to be in the negative digits. | time by VeriSign upon the USG's request. Such revocation or refusal to | issue a new cert after the first year of operation will leave the webserver | operator with a server that is no longer able to encrypt communications to | their customers in any meaningful way, thereby effectively shutting down | Internet based operations of the company unfortunate enough to invest in | such a flawed solution. I'll save Sameer the trouble, and suggest that people buy Stronghold. Adam -- He has erected a multitude of new offices, and sent hither swarms of officers to harrass our people, and eat out their substance.
At 09:16 AM 7/17/97 -0700, Bill Frantz wrote:
It seems to me that someone who has a one year export approved Verisign cert should use it to authenticate a new top-level CA cert which they pass to their customers. Cut Verisign and their nosy/noisy partner out of the loop.
Only a valid VeriSign Global ID cert (an X.509 v3 cert with a special extension) will activate the strong encryption in exportable browsers. This is hardcoded into Navigator and Internet Explorer. --Lucky Green <shamrock@netcom.com> PGP encrypted mail preferred. DES is dead! Please join in breaking RC5-56. http://rc5.distributed.net/
-----BEGIN PGP SIGNED MESSAGE----- In <3.0.2.32.19970717100056.00733db4@netcom10.netcom.com>, on 07/17/97 at 10:00 AM, Lucky Green <shamrock@netcom.com> said:
At 09:16 AM 7/17/97 -0700, Bill Frantz wrote:
It seems to me that someone who has a one year export approved Verisign cert should use it to authenticate a new top-level CA cert which they pass to their customers. Cut Verisign and their nosy/noisy partner out of the loop.
Only a valid VeriSign Global ID cert (an X.509 v3 cert with a special extension) will activate the strong encryption in exportable browsers. This is hardcoded into Navigator and Internet Explorer.
Yep, fairly simmilar to "policy tokens" that were disscused on the list last year. Let's face it people the management of Nut$cape & Mick$loth are a bunch of rat bastards who would sell their own mothers to make a buck. - -- - --------------------------------------------------------------- William H. Geiger III http://www.amaranth.com/~whgiii Geiger Consulting Cooking With Warp 4.0 Author of E-Secure - PGP Front End for MR/2 Ice PGP & MR/2 the only way for secure e-mail. OS/2 PGP 2.6.3a at: http://www.amaranth.com/~whgiii/pgpmr2.html - --------------------------------------------------------------- -----BEGIN PGP SIGNATURE----- Version: 2.6.3a Charset: cp850 Comment: Registered_User_E-Secure_v1.1b1_ES000000 iQCVAwUBM85khI9Co1n+aLhhAQFjlQP/a5Cp4KTDNKh0wzvF/Y6YoJOd0u2swbvv j02akFcab+mgfYGvEq9qUOmEoxw/jJ+4XXas0Zw/Ap6I8QcNEqRUQ/vAzR0UgUzo m039NpNzT8KTo8TtX6Ry2Zhob2Wk3cpZ2+1H+mOpmYKBbXd6PZfna4u6sk2W6BTk zNjh79taSDQ= =TuK9 -----END PGP SIGNATURE-----
Only a valid VeriSign Global ID cert (an X.509 v3 cert with a special extension) will activate the strong encryption in exportable browsers. This is hardcoded into Navigator and Internet Explorer.
So, write a new browser (Theres a linux project staring to do this for linux soon/now (i think anyway)) and then port to a bunch of systems and base it outside the US. There aren't any import restriction on strong crypto yet are there ?? Besides you could then set up a home page ala. netscape.com and rake in the advertising dollars, if you can get teh browser to take off. A lot of if's and work but in theory it works. Jason =8-]
It seems to me that someone who has a one year export approved Verisign cert should use it to authenticate a new top-level CA cert which they pass to their customers. Cut Verisign and their nosy/noisy partner out of the loop. ------------------------------------------------------------------------- Bill Frantz | The Internet was designed | Periwinkle -- Consulting (408)356-8506 | to protect the free world | 16345 Englewood Ave. frantz@netcom.com | from hostile governments. | Los Gatos, CA 95032, USA
At 10:00 AM -0700 7/17/97, Lucky Green wrote:
At 09:16 AM 7/17/97 -0700, Bill Frantz wrote:
It seems to me that someone who has a one year export approved Verisign cert should use it to authenticate a new top-level CA cert which they pass to their customers. Cut Verisign and their nosy/noisy partner out of the loop.
Only a valid VeriSign Global ID cert (an X.509 v3 cert with a special extension) will activate the strong encryption in exportable browsers. This is hardcoded into Navigator and Internet Explorer.
That's what patch installers are for ;-) --Steve
Bill Frantz writes:
It seems to me that someone who has a one year export approved Verisign cert should use it to authenticate a new top-level CA cert which they pass to their customers. Cut Verisign and their nosy/noisy partner out of the loop.
My understanding is that Verisign's licensing agreement explicitly forbids using any certs they issue as CA certificates. Maybe if the 'someone' paid Verisign an appropriate fee they might allow it, but I'd bet that fee would be very high. Verisign's no dummy, they don't want to enable new competition to ride on their backs. In the case of this special strong-crypto-allowing cert, Verisign would probably be encouraged to discourage cert holders from using the special Verisign certs as CA certs, for the very reason you suggest. :-) The format of the X.509 extensions that will enable strong crypto operation will be known soon. Even if Netscape et. al. tried to keep them secret, since they're public certificates they'll be available to anyone with an ASN.1 parser. -- Eric Murray ericm@lne.com Security and cryptography applications consulting. PGP keyid:E03F65E5 fingerprint:50 B0 A2 4C 7D 86 FC 03 92 E8 AC E6 7E 27 29 AF
Lucky Green wrote:
The cert is typically valid for a year, but is subject to revocation at any time by VeriSign upon the USG's request. Such revocation or refusal to issue a new cert after the first year of operation will leave the webserver operator with a server that is no longer able to encrypt communications to their customers in any meaningful way, thereby effectively shutting down Internet based operations of the company unfortunate enough to invest in such a flawed solution.
I don't know the details of the agreement between VeriSign and the USG. I'm curious: how will the CRL for this revocation get distributed? Since Communicator doesn't automatically pull CRLs, how can any action on VeriSign's part disable crypto for that server? Or are you suggesting that as part of the revocation process, the USG will bust down their doors and grab all copies of their private keys? -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
On Thu, 17 Jul 1997, Tom Weinstein wrote:
I don't know the details of the agreement between VeriSign and the USG. I'm curious: how will the CRL for this revocation get distributed? Since Communicator doesn't automatically pull CRLs, how can any action on VeriSign's part disable crypto for that server? Or are you suggesting that as part of the revocation process, the USG will bust down their doors and grab all copies of their private keys?
[Tom, I am glad that your are adding your voice to this tread]. It is true that Communicator does not presently pull CRL's. However, an X.509 based application probably should pull the CRL, or at least verify that a cert about to be relied upon has not in fact been revoked by looking for a match in the CRL. It stands to reason that Communicator will at one point add this, IMHO proper, feature. I also would like to mention the reader that yesterday's release of MSIE 4.0b2 *does* have the ability to check CRL's. Even if Communicator would never check CRL's, not even in the future, the mere fact that the Global ID cert have only a one year lifetime means anyone relying on Global ID can be held hostage by threatening to refuse to renew their cert. The reader may not be aware that unlike other certs, the Global ID certs are *only* issued by VeriSign. You can not go to a non-US CA and obtain such a cert. [Which of course would defy the whole purpose of this rather slick deal :-] Unless VeriSign includes in the price of the Global ID cert a bond that will compensate the buyer of a Global ID based commerce system for any and all future losses caused by VeriSign either revoking or refusing to renew a cert (fat chance), anyone basing their strategy on having such a cert is at risk of losing their business. --Lucky
Lucky Green wrote:
On Thu, 17 Jul 1997, Tom Weinstein wrote:
I don't know the details of the agreement between VeriSign and the USG. I'm curious: how will the CRL for this revocation get distributed? Since Communicator doesn't automatically pull CRLs, how can any action on VeriSign's part disable crypto for that server? Or are you suggesting that as part of the revocation process, the USG will bust down their doors and grab all copies of their private keys?
[Tom, I am glad that your are adding your voice to this tread].
It is true that Communicator does not presently pull CRL's. However, an X.509 based application probably should pull the CRL, or at least verify that a cert about to be relied upon has not in fact been revoked by looking for a match in the CRL. It stands to reason that Communicator will at one point add this, IMHO proper, feature.
I also would like to mention the reader that yesterday's release of MSIE 4.0b2 *does* have the ability to check CRL's.
Yes, we will add this feature in some future release. It will be configurable, so if the user doesn't want to check CRLs he doesn't have to.
Even if Communicator would never check CRL's, not even in the future, the mere fact that the Global ID cert have only a one year lifetime means anyone relying on Global ID can be held hostage by threatening to refuse to renew their cert. The reader may not be aware that unlike other certs, the Global ID certs are *only* issued by VeriSign. You can not go to a non-US CA and obtain such a cert. [Which of course would defy the whole purpose of this rather slick deal :-]
Aren't all certs VeriSign issues only valid for one year? This isn't any different. There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
Unless VeriSign includes in the price of the Global ID cert a bond that will compensate the buyer of a Global ID based commerce system for any and all future losses caused by VeriSign either revoking or refusing to renew a cert (fat chance), anyone basing their strategy on having such a cert is at risk of losing their business.
I fail to see the problem. Right now, if you want to communicate securely with exportable web browsers, this is the only way to do it. Either you do it, or you don't. If VeriSign doesn't renew your cert, then you're right back where you were the previous year. -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
Tom Weinstein wrote: | > Unless VeriSign includes in the price of the Global ID cert a bond | > that will compensate the buyer of a Global ID based commerce system | > for any and all future losses caused by VeriSign either revoking or | > refusing to renew a cert (fat chance), anyone basing their strategy on | > having such a cert is at risk of losing their business. | | I fail to see the problem. Right now, if you want to communicate | securely with exportable web browsers, this is the only way to do it. | Either you do it, or you don't. If VeriSign doesn't renew your cert, | then you're right back where you were the previous year. Nope, you've now got thousands of upset customers who were using secure communications, and are now using cheesy exportable ciphers. (Hopefully, you wrote CGI so you can quickly switch to using Stronghold. :) Adam -- He has erected a multitude of new offices, and sent hither swarms of officers to harrass our people, and eat out their substance.
Adam Shostack wrote:
Nope, you've now got thousands of upset customers who were using secure communications, and are now using cheesy exportable ciphers. (Hopefully, you wrote CGI so you can quickly switch to using Stronghold. :)
It wouldn't help. The problem is that your customers (using the export client) who used to be able to connect using strong crypto now have to use weak crypto. It doesn't affect people using non-exportable clients. -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
On Thu, 17 Jul 1997, Tom Weinstein wrote:
Lucky Green wrote: Even if Communicator would never check CRL's, not even in the future, the mere fact that the Global ID cert have only a one year lifetime means anyone relying on Global ID can be held hostage by threatening to refuse to renew their cert. The reader may not be aware that unlike other certs, the Global ID certs are *only* issued by VeriSign. You can not go to a non-US CA and obtain such a cert. [Which of course would defy the whole purpose of this rather slick deal :-]
Aren't all certs VeriSign issues only valid for one year? This isn't any different.
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
There's probably no technical reason these patches must originate with Netscape. Seems like a healthy cottage industry could spring up to supply patch software to offshore companies which want magic certs w/o USG approval. --Steve
=snip=
I also would like to mention the reader that yesterday's release of MSIE 4.0b2 *does* have the ability to check CRL's.
Yes, we will add this feature in some future release. It will be configurable, so if the user doesn't want to check CRLs he doesn't have to.
An excellent way of implementing it.
Aren't all certs VeriSign issues only valid for one year? This isn't any different.
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
uh, why does one need permission of the usg to issue "magic certs"? ------------------------ Name: amp E-mail: amp@pobox.com Date: 07/18/97 Time: 04:06:36 Visit me at http://www.pobox.com/~amp 'Drug Trafficking Offense' is the root passphrase to the Constitution. Have you seen http://www.public-action.com/SkyWriter/WacoMuseum ------------------------
amp@pobox.com wrote:
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
uh, why does one need permission of the usg to issue "magic certs"?
Because issuing these certs is defined as a "defense service". -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
At 10:02 PM -0700 7/18/97, Tom Weinstein wrote:
amp@pobox.com wrote:
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
uh, why does one need permission of the usg to issue "magic certs"?
Because issuing these certs is defined as a "defense service".
Precisely why there needs to be some hacks, from offshore CP, to enable anyone to issue "magic certs". Any volunteers? --Steve
Steve Schear wrote:
At 10:02 PM -0700 7/18/97, Tom Weinstein wrote:
amp@pobox.com wrote:
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
uh, why does one need permission of the usg to issue "magic certs"?
Because issuing these certs is defined as a "defense service".
Precisely why there needs to be some hacks, from offshore CP, to enable anyone to issue "magic certs". Any volunteers?
The main result of such a hack would probably be to get our export license yanked. Of course, you might see that as desirable, for either its direct or indirect effects. -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
-----BEGIN PGP SIGNED MESSAGE----- In article <33D04A75.31E08282@netscape.com>, Tom Weinstein <tomw@netscape.com> wrote:
amp@pobox.com wrote:
There's nothing preventing another CA from getting permission from the USG to issue these magic certs. We would have to distribute a patch, but I don't see any problem with that.
uh, why does one need permission of the usg to issue "magic certs"?
Because issuing these certs is defined as a "defense service".
It is in no way a defense service for Ian's Certificate Authority to issue a digital certificate to Steve's Offshore Laundry, Inc. that basically says "I think communications to the holder of this cert should use 128-bit encryption.", even if it uses the same V3 extension that Verisign uses. Now, if some company were to sell a browser overseas that enabled 128-bit encryption when it saw _any_ cert with this extension (or even any such cert from a CA in the user's trusted CAs list), I'd say it's the browser company that's supplying the encryption, not the CA; the CA just issued a signed statement of fact/opinion. It would seem to me, though, that the only reason Netscape was able to release a browser with the "128-bit-if-Verisign-magic" mode overseas was that the USG had gotten Verisign to agree that it wouldn't issue Verisign-magic certs to "alledged terrorists", etc. If Verisign renegs on the agreement, and issues the Verisign-magic certs to left-handed albino money-laundering aliens, they'd be in violation of whatever they signed with the USG, but certainly not in violation of the crypto export regs (which, now that they're under Commerce, I'm not sure even have a "defense service" category anymore). So in answer to the original question (IMHO), you don't need the permission of the USG to issue "magic" certs (ones with the V3 extension). It's just that browser companies won't be allowed to make browsers that turn on strong encryption for _your_ "magic" certs unless the USG trusts you not to give such certs to just anybody. Contrasting this situation with Microsoft signing CAPI modules is left as an exercise for the reader. - Ian "I believe that the bearer of this signed message should be entitled to use as strong crypto as he likes." -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBM9FR/kZRiTErSPb1AQG0ogP9HC1bMyak7D1PEgRHVHPYU+a5BzTpyf/W 4aYINON+eKxw0PbDM6Q6FjnP8r1dXSBPH1T8v+2RbTqQ0A4bGVEZWGlcJv5jzuRG pJb/PuZQwNgecp2sx/sniyfHJdhE6H4omiaDa2URO00Mr9s7iotFleC5LdgGg+XV n9EeJJDxLtY= =mp59 -----END PGP SIGNATURE-----
I think it tells us that Verisign managed to convince the government that their product is only used for authentication, not encrypting content. Which appears currently to be true, no? And since AFIK (Please, someone, correct this if I'm wrong!) you can't with netscape anyway download another party's key that you verify with a Verisign certificate, it would take a fair amount of work for the ordinary user to set up a secure channel using the current Verisign infrastructure. The ITAR exception for authentication-only products is of long standing. On Wed, 16 Jul 1997, Bill Stewart wrote:
Forwarded from PGP-USERS list:
First PGPInc and now VeriSign? Hmmm. Is this telling us something?
"VeriSign on Monday said it received permission from the U.S. Department of Commerce to export 128-bit strong encryption software and issue digital identifications to approved organizations based on that software. "
"Under the 128-bit scheme approved by the U.S. government Monday, companies will not need to place their encryption keys in escrow, or submit to U.S. government key-recovery requirements in order to use VeriSign's software, company officials said."
# Thanks; Bill # Bill Stewart, +1-415-442-2215 stewarts@ix.netcom.com # You can get PGP outside the US at ftp.ox.ac.uk/pub/crypto/pgp # (If this is a mailing list or news, please Cc: me on replies. Thanks.)
A. Michael Froomkin | +1 (305) 284-4285; +1 (305) 284-6506 (fax) Associate Professor of Law | "Cyberspace" is not a place. U. Miami School of Law | froomkin@law.miami.edu P.O. Box 248087 | http://www.law.miami.edu/~froomkin Coral Gables, FL 33124 USA | It's @%#$%$# hot here.
On Thu, 17 Jul 1997, Michael Froomkin - U.Miami School of Law wrote:
I think it tells us that Verisign managed to convince the government that their product is only used for authentication, not encrypting content. Which appears currently to be true, no? And since AFIK (Please, someone, correct this if I'm wrong!) you can't with netscape anyway download another party's key that you verify with a Verisign certificate, it would take a fair amount of work for the ordinary user to set up a secure channel using the current Verisign infrastructure.
True, the certs themselves are not covered by the export controls. But we aren't talking about export law. We are talking about a four way contract between Netscape, Microsoft, VeriSign, and the US government. Under that contract: o VeriSign will only issue Global ID certs to US companies with all their servers located in the US and overseas banks with servers abroad that play by the USG's rules. Once the USG no longer approves of the participants using strong crypto with their customers, VeriSign will revoke the cert, disabling secure communications, and thereby severely damaging, if not destroying, the business of the party unfortunate enough to have relied on such a cert for their livelyhood. o Netscape and Microsoft get a blanket approval to ship their servers to non-US banks that meet the USG's criteria. o Netscape and Microsoft also receive approval to export browsers that can use strong crypto *exclusively* with sites the USG and VeriSign approve of. o The USG no longer has to waste time handling export applications it doesn't mind approving anyway, such as those for US-friendly foreign banks. And the USG no longer has to listen to US companies complain because they are unable to provide their non-US customers with secure access to the sever located in the US. Lastly, and most importantly, every purchaser of a VeriSign Global ID cert allows the USG and VeriSign to install a MASTER-OFF switch in the heart of their business. I feel sorry for the poor suckers that will lose home and hearth after subscribing to this fatally flawed solution. --Lucky
Bill Stewart wrote:
Forwarded from PGP-USERS list:
First PGPInc and now VeriSign? Hmmm. Is this telling us something?
"VeriSign on Monday said it received permission from the U.S. Department of Commerce to export 128-bit strong encryption software and issue digital identifications to approved organizations based on that software. "
"Under the 128-bit scheme approved by the U.S. government Monday, companies will not need to place their encryption keys in escrow, or submit to U.S. government key-recovery requirements in order to use VeriSign's software, company officials said."
What this means is that VeriSign is now allowed to issue the "magic banking certs" for servers that allow them to communicate with the export version of Communicator using 128-bit SSL. On the other hand, it's also meaningless until "approved organizations" start getting approved. -- What is appropriate for the master is not appropriate| Tom Weinstein for the novice. You must understand Tao before | tomw@netscape.com transcending structure. -- The Tao of Programming |
participants (13)
-
Adam Shostack -
amp＠pobox.com -
Bill Frantz -
Bill Stewart -
Eric Murray -
iang＠cs.berkeley.edu -
Jason William RENNIE -
Lucky Green -
Michael Froomkin - U.Miami School of Law -
Steve Schear -
Tom Weinstein -
tomw＠netscape.com
-
William H. Geiger III