|
This file is available on a Cryptome DVD offered by Cryptome. Donate $25 for a DVD of the Cryptome 10-year archives of 35,000 files from June 1996 to June 2006 (~3.5 GB). Click Paypal or mail check/MO made out to John Young, 251 West 89th Street, New York, NY 10024. Archives include all files of cryptome.org, cryptome2.org, jya.com, cartome.org, eyeball-series.org and iraq-kill-maim.org. Cryptome offers with the Cryptome DVD an INSCOM DVD of about 18,000 pages of counter-intelligence dossiers declassified by the US Army Information and Security Command, dating from 1945 to 1985. No additional contribution required -- $25 for both. The DVDs will be sent anywhere worldwide without extra cost. | |||
19 December 1999: Revise Gladman address and URL.
17 December 1999: Add Hickson and Gladman messages.
16 December 1999
From: "Brian Gladman" <brian.gladman@btinternet.com>
To: <ukcrypto@maillist.ox.ac.uk>
Subject: US and UK Government Crypto Policies
Date: Thu, 16 Dec 1999 13:07:35 -0000
Over recent weeks I have had the chance to discuss the proposed developments in US crypto export controls with a number of people who have been quite closely involved and this has thown some new light on what is going on.
One area of confusion in the proposed new regulations was the introduction of a new category of cryptographic product called 'retail'. This gave me a problem in interpreting the changes since we already had 'commodity', 'mass market' and several other designations so it was not obvious why yet one more category was needed. I felt that it possibly meant 'cryptographic software used in support of e-commerce applications' but I am told that this is not the case.
The US industry view, claimed also to be a US administration view as well, is that 'retail cryprographic software' is any cryptographic software that is (or could reasonably be) sold to the public through retail outlets. This term is seen to include mass market and commodity items but in addition items that are not mass market or commodity but are stil sold through retail channels to consumers (for example, low volume products for specialist applications).
I am also told that the anomaly of being able to freely export 'open souce' cryptographic software in source code form but not in binary is recognised as silly and is being changed so that neither source nor binary is controlled.
Several US industry people who I trust are very commited to the view that this IS a radical change in the direction of US crypto export control policy that spells the end of crypto export controls on everything expect high grade 'military' and 'government' crypto. I am not so sure as they seem to be about this because US industry has been let down by the administration many times on such matters but I have to say that I have never heard such very high expectations so consistently expressed in previous rounds.
But since all 'software only' cryptography is essentially flawed for serious use, I have been asking my US colleagues how these changes might impact on hardware deliverables. This gets even more interesting since more than one significant US crypto hardware supplier claims that they will be free to deliver unrestricted strength crypto hardware under the new regime without the need for licences or product review. I am very suprised by this claim, and somewhat disinclined to believe it but if it turns out to be true, then the changes are significant.
The area of cryptographic interfaces remains a major problem area. For reasons that people in US industry don't fully understand (or cannot explain or are not allowed to explain), products offering 'crypto shaped holes' are problematic even though crypto thats fits into such holes may be freely exportable!
For example I have been told by more than one US company that they will be able to deliver hardware and software solutions that implement high grade retail cryprographic products with a '3 call API' - set key, encrypt and decrypt - without the need for export licenses. But it seems that Microsoft will not be able to deliver an OS with a hole that can accommodate such products without a license. This seems completely odd and looks like an attempt to break the Microsoft stranglehold on the OS market since Linux will be available with crypto shaped holes for which US crypto companies will be able to provide high grade crypto solutions.
All of this could be seen as a recognition that crypto for personal privacy is now unstoppable but that wider business use can still be managed by controlling its availability on high grade servers and the like. But this is hard to believe since this would only work to further strengthen Linux and make it the dominant force in the server business at the expense of Microsoft and other US suppliers.
In overall terms my suspicion is that this is one of several areas where US authorities have not thought through the full consequences of changes that have been proposed. Hence the delay while many new scenarios and unthought of are worked through. I suspect the delays in this new set of regulations were hence inevitable.
Rightly or wrongly much of US industry is very clearly expecting most (or even all) controls on crypto for non-miltary, non-government use (expect for 'nasty' countries) to be removed. There is also a surprisingly high degree of confidence that this is what is on offer. But while some in the US administration may want to deliver just this, I suspect that others did not see these changes as anywhere near this radical.
With such very, very high expectations in US industry, the backlash if the
US administration fails once again to deliver will be truly extreme.
And with the political processes still in place and running in case they
do fail, I doubt that the US administration can risk non delivery this time
round.
While it is US policy rather than that of the UK that really matters (since the latter will simply follow that of its US master on such matters), it remains a sad reflection of the UK government that, despite all its rhetoric on e-commerce, it has steadfastly ignored calls for changes in its crypto export control regime and has even sought, in the PIU report, to present its actions in a highly misleading way.
But the UK public is now becoming aware that the crypto algorithms used by the mobile phone companies and network operators are deeply flawed to the point that at least one of them can be broken in real time on a home PC!
In Europe these standards are agreed by a body known as the European Telecommunications Standards Institute (ETSI), a grouping that involves both government and telecoms industry representation but which does not appear to be subject to any democratic scrutiny or control. It is now clear that UK government and UK telecoms industry groupings have been actively conspiring within ETSI to consciously and deliberately undermine the privacy of UK (and European) citizens. Moreover at least one of the companies involved in this is now suggesting that UK citizens should actually rely on it for 'trust services' in support of e-commerce!!!!
It is hence clear that there has been a conspiracy between government and industry in the UK and in Europe to deliberately undermine the privacy of citizens. Its all very well for the EU Parliament to get 'up tight' about Echelon and the NSA but it would do well to look at what is happening in its own back yard!
So we can now see that the UK government has been at the forefront of actions in ETSI to ensure that no effective cryptographic protection was ever introduced for mobile phones. Worse still, major UK based companies have acquiesced and even cooperated in this. And the UK Government wants UK citizens to feel comfortable with a recommendation that such matters should in future be considered through government/industry consultation with no public interest involvement.
A letter from CR&CL(UK) direct to the UK Prime Minister expressing serious concerns about these matters has been treated with complete contempt, with not even the common courtesy of an acknowledgement.
Moreover, as might be expected, related letters to IBM and British Telecom have produced very different results.
IBM has responded and has given assurances that it is aware that very serious issues are involved. British Telecom, on the other hand, like the UK Government, has simply sought to ignore the issue completely, presumably because it is hopelessly implicated alongside the UK government in actively working to undermine the privacy interests of UK citizens.
There is hence still some hope that the computer industry will prove to be 'good guys' but relying on the telecoms industry for privacy (or trust services) is like dining with the devil in the hope of salvation.
Brian Gladman
Mr. Gladman's Web site: http://www.btinternet.com/~brian.gladman/
From: "Brian Gladman"
<brian.gladman@btinternet.com>
To: <ukcrypto@maillist.ox.ac.uk>
Subject: Re: US and UK Government Crypto Policies
Date: Fri, 17 Dec 1999 12:01:11 -0000
>From: Nigel Hickson <nigelhickson@compuserve.com>
>To: <ukcrypto@maillist.ox.ac.uk>
>Sent: Friday, December 17, 1999 8:56 AM
>Subject: US and UK Government Crypto Policies
>Brian and colleagues
>Good stuff. But as you well know we do not simply follow US on export
>controls on crypto; we are tied into the EU Dual Use Goods Regulation.
>And indeed we have made considerable moves; what other EU countries have
an
>open general licence for all "retail" software?
>Nigel
Thank you for your comments Nigel. I hear what you say but I am commited to the view that the only way to read crypto policy is to ignore the words and look at actions instead. I am pleased to hear about the OGL for all retail software but this is the wrong way round - that is "we retain control but we give you permission to..." - I wan't to see all crypto controls REMOVED and a UK government policy that is HONESTLY the same as that of Germany.
Both the UK and Germany are equally 'tied' to EU Dual Use Goods Regulations but once we look at actions rather than words, we can immediately see that their crypto policies are completely different. Anyone who is serious about crypto policy knows only too well that the EU regulations in respect of cryptography are so vague that virtually any interpretation can be put on them. Hence it is equally possible for both Germany and the UK to claim compliance while one promotes the use of strong cryptography and the other works vigorously behind the scenes to undermine it.
If we look at the words of Germany and the UK, we have seen a very clear commitment from Germany earlier this year to promote the deployment of strong cryptography. When this was announced your own comment was that it was nice to see that Germany and the UK had such similar policies. I am sure people on this list will remember this well.
But looking instead at actions, we find that Germany is contributing to the funding of GPG while your own organisation - the DTI no less - has been leading work in ETSI to ensure that any cryptography that is deployed in mobile phones is so weak that some of it can even be broken on home PCs! Moreover, we have a UK government (or more likely a civil service) that is prepared to lie to the UK public about its actions in these areas as it did in the PIU report (call it being economical with the truth if you like).
The reality in the UK is that the GCHQ axis still dominates the formulation of UK crypto policy. The GCHQ management (and its civil service lackies) will do literally anything that is necessary to meet the wishes of their US masters with ***complete disregard*** for any impact that this will have on UK citizens or on our European partners.
The mobile phone crypto saga makes this point very well. The crypto in mobile phones protects only the 'over the air' segment and this means that traffic is in the clear at the base stations and in transit within the fixed telecoms network. As a consequence law enforcement needs can very easily be met without the need to interfere with the 'over the air' segment. And since all responsible nations want to work to prevent crime and terrorism, they will be prepared to cooperate in the exchange of such information. There is hence no strong 'crime and terrorism' case for making the 'over the air' segment of mobile phone systems weak.
But while the US and the UK can get at their own traffic at base stations, they cannot get at German and French traffic in this way and this means that the ability to spy on these and other nations does depend on ensuring that weak algorithms are deployed. In consequence, in order that NSA could continue to spy on Europe, it was vital that a poor algorithm should be deployed in EU mobile phone systems. Hence the UK was instructed by its US masters to ensure that EU deployed mobile phone systems were unprotected. However GCHQ could hardly go into ETSI and do this for itself so it simply turned to its UK civil service sympathisers to do this - hence the DTI was sent in to do GCHQ's dirty work for them.
And the biggest surprise of all is that our EU partners have fallen for this US/UK Trojan Horse for so long. Echelon is one thing but the subversion of ETSI by the UK to meet US ends is in an entirely different league. The European Parliament needs to look at its own backyard!
And we should also note that major industry sectors have acquiesced in undermining the privacy interests of EU citizens. And the FEI has just published proposals for the voluntary control of trust services provision which are heavy on industry involvement and lightweight in the extreme on public interest involvement and oversight. Lets hope that they see the critical weakness in this approach before its too late.
Brian