Turing and origins of UKUSA

[Caspar Bowden (travelling)]

On Turing Day +1 thought ukcrypto might enjoy this....

As list members no doubt recall al couple of years ago the UK National 
Archives and the NSA simultaneously published a lot of material on the 
UKUSA intelligence sharing agreements originating in WW2. However the 
NSA published <http://www.nsa.gov/public_info/declass/ukusa.shtml> 
significantly more (and different) material than released in UK 
<http://www.nationalarchives.gov.uk/ukusa/>, and I was intrigued by 
several aspects of the US "early papers 1940-1944" 
<http://www.nsa.gov/public_info/_files/ukusa/early_papers_1940-1944.pdf>

Turing visited the US in November 1942, mainly to inspect US production 
of bombes and have a shufti at US methods, but also to look at work in 
Bell Laboratories on a new speech scrambler (likely what became SIGSALY 
<http://www.nsa.gov/about/cryptologic_heritage/center_crypt_history/publications/sigsaly_story.shtml>). 
However he was refused permission, and the "early papers" document the 
US Army side of an escalating row which lasted until a prototype of 
UKUSA was concluded in May 1943 (long before BRUSA in 1946)

The row was about the fact the US had become suspicious the UK was 
holding back info on the Lorenz machine cipher (Tunny), although the US 
had briefed the UK on the breaking of the Japanese PURPLE; also that the 
UK wanted to keep control of Enigma exploitation because of worries 
about security (reasonable because the US wouldn't tell them the 
technology the US wanted to use to protect the dissemination of 
decrypts); and that also the US Navy had got full access to UK decrypts 
of German U-Boat Enigma but such agreements hadn't been reached with the 
US Army for the European of African theaters.

Previous primary sources include Turing's initial report 
<http://www.turing.org.uk/sources/washington.html> (Nov 28th)  of his US 
trip (released in 2004), which opened

I reached New York on Friday November 12th. I was all but kept on Ellis 
Island by the Immigration Authorities who were very snooty about my 
carrying no orders and no evidence to connect me with the F.O. They 
considered my official's passport insufficient in itself. They asked me 
very minute details about where I was to report etc. I think it might 
have been better from a security point of view if I had been provided 
with some kind of document of the kind they wanted, to say nothing of 
the possibility that I might have been held until Stevens or somebody 
identified me

..and continues with understated humour about the US approach to the 
work. Turing is optimistic in the report that "all now seems to be well" 
re: problem with visiting Bell Labs, but the "early papers" show that he 
did not get permission until Jan 9th. His UK minder Maj.Stevens in a 
covering note supports Turing skepticism and adds
. T
They (the US) are jokingly credited with wanting to take all traffic 
that comes in and subject it immediately to every known process, 
regardless that some of it may be P/L or in a cipher which they hold.

Amazingly Turing had not had instructions about whether he was allowed 
to brief the US on Tunny (i.e. that by this time Tutte had reconstructed 
the Lorenz machine purely with manual analysis), and evidently had to 
keep this from the formidable US cryptographer Friedman (that must be 
one of the all time cagey conversations)

What I haven't seen written up in any historical work since 2010 is that 
relations became so bad in early 1943 that the UK were contemplating 
cutting off the US from Continental Enigma (at least the US Army thought 
so, and advised to call what they assumed was a UK bluff). The 
corresponding documents on the UK side weren't released. The US Army 
resented the fact they got trumped "in the Turing case" and that GCCS 
had access to Churchill "and therefore to F.D.R" which they evidently 
lacked.

There are several gems in the documents but favourite so far is :

"They (the UK) set forth the claim that in connection with this whole 
subject of secret communications equipment, either voice scrambling, 
cipher machines or anything of a similar nature, the specialists who are 
experts in cryptanalysis or descrambling, should be in on the initial 
development of the equipment. In that way these experts (according to 
Tiltman et al) can point out weaknesses in design which could be 
corrected in the development period. They claim that hundreds of man 
hours could be saved if this procedure were followed rather than to have 
a machine developed in one laboratory and then to give to another 
laboratory the job of breaking down its traffic. In my opinion, this is 
merely another attempt to gain access to technical information on our 
cipher machines and ultra secret scrambling devices and is not a 
plausible argument" (Dec 17 1942)

CB


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.chiark.greenend.org.uk/pipermail/ukcrypto/attachments/20120624/ee352c07/attachment.html>