Historical Document · 1950
Special Conference on M-209 Security — minutes, 12 October 1950
Minutes of a 1950 meeting on the M-209's security and on German cryptanalytic work against M-209 traffic in WWII.
Historical Document · 1950
Minutes of a 1950 meeting on the M-209's security and on German cryptanalytic work against M-209 traffic in WWII.
• SPECIAL C<m'ERENCE <A"'f ll-209 SmtJRlTY At&ached are minutes ot a meeting held. 12 October 1950 at 1400 hours at the Naval Security Station, Building 19. Room 2.30 1 to discuss general aspects ot the securit7 of the American cipher .machine li-209, and German crypt~1c work on traffic enciphered by this ms.chine during World War !Io Declassified and approved for release by NSA on 06-10-2014 pursuantto E .0. 1352a TOP SECRET .. SPEDIAL CONFEP~NCE at J.lf-209 SJOOURITY Jlinutes ot meeting held 12 October 1950 at 1400 hours at the Naval Securit.y Stat.ion, Building 19,, Room 2300 At the invitation of Jlr. William Fo Friedman, Technical Coneultant to the Director1 AFSA,, those present were: llr o T. Ro Wo Burton Miller British representative .llro Kenneth Lo Perrin British representative Dr o Abraham SinkDv 1 AFSA...04T Mro Frank Austin, AFSA....41 Dro &.. Wo Pettengill, AFSA-14 Captain Uar7 Co Lane,, AFSAc...14 (Recorder) gpeaj.ng Remsrks by Kr o Friedman:: Jlr o Friedman opened the meeting b,,y stating that the purpose of the gathering was to discUBs in general the security ot the .American cipher machine H-~9, and in particular the amount of K-239 traffic solved by the Germans during \iorld Wa.r II and t.he amount. of int.elllgence they obtained therctrom.o Mr. Friedman said that the question had arisen during a recent discu8s1on in which the British repreaentative 1 Mro Killer• had referred to the 11 vast amount or intelligence ~hich had been obtained by the Germans from the reading or tratfio enc17Pted "With the American cipher machine 14=209 11 11 a statement the validit.J" of which Hr o Friedman had questioned and \1h.ich ce had asked Ur o Hiller to discuss with himo llro Jliller had in turn asked Jlro Perrin to gather information on the matter trom TlCOU aourceeo American representatives from. AFSA='J4 and AFSA...14 had been invited by Hr o Friedman to join in the discussiono llro Fried.man added that., although the question was largely academic~ the 1!-209 being noJ1=.0perational, he believed nevertheless that the discussion would prove of interest and benefit to the participe.ntso Remarks by Ul·~..Jlill..!!,: Kro lliller remarked that. tile discussion might. tiell have a certain amount o.f operational value, since he understood. that at the present time the H~209 .machine was being used for communications b7 the U.So Arlfll" in Korea~ With the participation or British forces in Korea, the question had arisen or a coll!ll!On cryptographic SY"Stem t.o be employed by combat unit.so Whereas the British are using Typex and would be prepared to u.se t.he CCM at Brigade level,.the 11....209 is at present. the only system. available to U.So forces at the equivalent U.So levelo Unless the U..S., Arnv were prepared t.o put the CCK rurther forward, therefore, it appeared that there might ba a possibility of having to use the H--209 as a combined s7stemc. A discussion of its security was thus a matter of considerable interest to both parties at the present timeo This statem~nt was corroboraLed by llro Austino Mro Miller then introduced M.rr. Perrin, who had prepared remarks on the solution of K=209 traffic as noted in TICOK sources~ Remarks bz_ ifro Perrin: Mro Perrin expl.s.inE'd that he ha.d divided his statement into four parts: Ao Hol1 much M-209 traffic ~as read by the C-erma.ns; B. How much time \1as consumed in the reading or t.he traffic,; Co How v.a.s the traffic read; and 1 Do What were tbe intelligence results or the reading or the traffic .• 2 TOP SECRET • Ao flow lilUCh K-209 t.raf'tic was road bz: the Ger.maria: §.latsmeqt bz llro P!11•r1n: VariOtJs .figures were given by the Ge~s uho 1u1.; e int.enogat.eei concerning t.ba amount. ot Jt-209 traffic Vfhich ttas re~ • .llro Perrin attir.med hie be.lief,, hOYteV'er,, t.hat. these diacropanc1es retlected actual dirrerencea and t.nat 'the reports,, 'While vacying, were true, since apparenti,- t.he armunt. or t,raftic solved differed according to the originating serviceo Thus, tor inst.anoo, almost. no U.S .. Naval 11-209 traffic was read: firatq, because t.here was lit.tle traffic, and second]3, because there •as a rolat1ve]3 st.rict cipher discipline in t.he.t serviaeo On the other hand.,, moat. ot t.he eources stated ttJat :trOlll. 6 "" 10% of u.s,, Army )[....209 trat.tic 'ftae reado One analyst. ot the German ArlB7 Cryptologic Agency, Rudol.t' Hentze• bad boasted that. 30i ot K-~ tirattic was reado llro Perrin admitted that Hentze ¥.as poasibl.¥ an unreliable sou.rceo A German ArllJY'of'ficer~ Lto Schubert,p had at.ated that ll....a'.>9 tratf'ic was read tor about. 20 da.rs a montho Although t.his might im.p4" that tw0=>th1rda of t.he total traffic was read.a llra Perrin stated that the tigure presumabq related to solution of occasional keya running concurrent.~ during the month rather than to the total amount ot daily traf'tic enciphered, and did not therefore imp]¥ that t~o~thirds or Lhe total trat.t'ic had been res.do Somewhat greater success was claimed by the senior cr1Pt.anal1'st. of t.he German Air Force, Voegele, who said that from 6 c:. 8 days ot .11-.209 traf'fic on the Western Front was read each mnth by hi& unit, and some l - 2 day's traffic in the Mediterranean areao Remarks bz: lira. Austi!,l: llra Aust-in pointed out that the solution of tor 20 days a month, provided that Schubert 9 s figure referred, as llro Perrin suggested, to the solution of occaeio.ne.l keys~ did not represent a large percentage of the total amo1111t ot Jl..,.Z)9 trattic pa.aaed 9 since at somettimes during t.he war as many as 7S dit.ferent ke.fS were being used simultaneously, each divisJ.on and corps being provided with its own set 0£ ~raffic ke7e,. ae were the numerous Milit.a1•7 Police battalions and units of the Army Air Forcea J&.ro Aust.in t.hen eta.t.ed that U.S.Ao monitoring of the M-209 traffic or an American Armored Division in England prior to D-=-Dq had resulted in the reading cr:ptan!J.,yticalJ.y ot 29 days' traff'ic a montho Alt.hough corrective meaeuree=had beei6 taken betora ths invasion. the heavy casualties a.mong U.So conuzwnica.tions personnel after the invasion he.cl made necessary the use or untrained Ar1Jf3 personnel as cr7ptographic clerks-. with a consequent decline in cipher disciplineo Overall estimates by Am3rica..~ experts on the amount of 11.~ro9 trat.ric probably read by- the Ger.mans during the war waa set. at io,g,, including physical compromisesa of which several cases 11.ere lmowno An estimate of 6% ia accepted £or the reading or ll~2>9 traffic by pure}3 ~ryptanalytic m.eanso Repk of l{rc. Perrin: Mro Perrin agreed in aubstanc~ ¥eit.h Ura Austin°a l'"emarks but set the reading or M.-209 traffic at 10% without including that resulting from phy'sical compromise. believing tnat the £1giire ~ith physical compromise innluded would be somewhat highero Bo How mch time waa cgnsumed in the re&diqg of t.he traffic 1 Statement. bz Mro Perrin: Mro Perrin stated that., alt.hough there was no information .trom TICCll sources a.bout the amount or ti.'D.e required for the reading ot messa• ·es in depth 6 it was presumed that this would take some two or three hours., There was. how~er, general agreement in TICal reports as to the time conswaed in b1•ealdng EJ!lative settings: this was usual:cy- aet at t"Wo or three hours (two hours \·.ere given by luziusj) three b.T Hent~e, two to three b.T Barthel. all German Army cryctanalyste) o Estimates on the amount of time required to obtain the abaolu.te se.ttin& varied: 12 hours was given by Hentze (Arm;y), lc.2 dqs by Barthel (Arl!\Y) ~ 22 hours by Voagele (Air Force), 3 TOP SECRET • 72 hours bl' Schult.ze (flavy). llro Perrin sums-est.ad that. 24 hours m.ight b& ta.ken as an average "it one "ere lucqno Thus• a total of approximate]¥ 28 hours would be needed atter the messa,;es "'ere received by the section charged with solution of the traf'fic, and it was general]¥ admitted by the German cr,ypt~ts that. the Jg...2)9 t.raftic was received several dqs ai'ter inte:iception (according to Voegele 7-B da,ys)o Remarks bz: at.hers: Jlro Friedman remarked that,, since cond1tions were usually unfavorable,, a t.ime lag or t.wo or three weeks between interception and the obtaining ot the absolute settings for the Jl...3>9 trai'.i'ic would not. be improbablea Mro Austin voictJd the opi."lion that the obtaining ot the absolute set ting would require at least 1.8 hours, or mo1•e often up to tour &Q"e, attar the messages "Wez·e broken in deptho In support of this view Jlro Austin noted that. the absolute setting had been obtained bT U.S. crn>tanaJ.7sts in 18 hours under the most favorable circumstances; Jlro Austin also cited the German Ar'llf/ cryptanalyst 11 Graupe,, who gave an estimate ot 5 ~ 7 days ae the tin:e necessary tor obtaining the absol.ute · setting. Captain Lane remarked that the C--er.man source quoted by Hro Austin was perhaps less reliable than those quoted by Uro Perrin,, a remrk with which Mr. Perrin a.greed. Gene:ral Agreement: It was genera.lJ.3" agreed that the German crypt... ana.]¥sts ~ere hampered b7 two factors or decisive importance in their solu.tion of Jl=.209 m.eseages. One was the lack of any crn>tana:cytic mechanical device tor tacilitating or expediting the solution, thus forcing all work to be done b,y hand; the other wan the time lag between interception and receipt of uassagee °b1" the cryptana.lyt.ic units in Berlino In reply to a quer7 .from Mro Friedman whether solution genera.lq took place at the front. or in the central unite, Yro Perrin said that frontal solution ot the )(....~9 was apparently confined to Lhe reading ot messages in depth,, the deter.mining o:t absolu~e settings being accomplished at the central uni&eo How was the traffic read: Oo · Remarks bz Kr" Perrin.; Mro Perrin etated that the German crypta.'lalyats never developed a general method of solution, although there is a description of a theoretical solution of the K=-209, given a text of 30<)0c.SOOO letters., For practical eolution ot the traffic German cryptan~sts depended either on depth or on operators' errors~ Cqrrobora.tion by llro ~ustin and Drc_P~j;.te~_!!: Mro Austin and Dr a Pettengill corroborated this stat.ement of Jlr o Perrin and agreed that the theoretical solution of tho M=.209 had been o:t no practical va.lueo Dr. Pett.engill added that the Germans had apparently' developed a m.eehanical met.hod of solutior., but the principles of this method were obecure_. and there was no conclusive evidence that it had actually been emp).oyedo Do l!bat were tqe intellig':!P.,.~~ rest!lts of the r~adin.$ of the ,trafi'i,9.: Remarks by Kr .. Perrin: Bro Perrin remarked l:.hat, because or the time lag between the in~e1'Ception ot messages and their receipt by the crypta.nal.ysts, the tactical intelligence gained from the reading or Jb..209 traf'fic by puroly cryptana.1¥f:.ic means was negligibleo Nevertheless, there was an appreciable amount ot strateg!E., intelligence which was gained from the reading of the .11.-209 traffic, such as t.b.e off'ect ot the German v....1 bomb_, the building of new air.fields) and Allied order of battleo Mr .. Perrin pointed out that, had the Germans been able to organise their own intercept and relrq services more etticientl,y9 the solution of .14...209 messages would have been of tactical. value,, As it was, he admitted that the value ~as main.Ly strategico 4 TOP SECRET ' Remarks by Ot.hera: Mr o Friedman hereupon interposed a question as to the readibility ot s7stem:! ueed by the British in comparable ecbelonso Mro Perrin stated that the British equivalent systemb the Stencil SUbtractor (SS) Frame, was never read bJ' the Germans except for ona month 'l'fhen the basic book had been capt.ureda Dr. Sinkov remarked that the British cipher had two dieedvantagee from. the American point ot vie\ilo One wae the cumbrous nature of tlle syat..em.; the other Yl&B the tact that Am.arican operators are not trained in sending or cop.Ying digits,, and thua.. tind a s,ystem like the SS Frame ditticW.t to usea Jlr" lliller added t.hat. the Stencil Subtractor F.rwne had also been less uaed than the American J&=.209, eince the British cipher machine i)pex was used e.t a lo~er echelon than t.he cOJRparable American device. and thus 'l)'pex relieved the stencil Subtractor Frame or some 400 ... 500 messages dail.70 llro Austin corroborated Mro U:l.lleroe remarks and stated that. t.he U.Sc ArJJJ¥ was so organized that the bulk 0£ regimental and combat level trattic was necessarily" encrypted with the )l...2090 Q.oncluding Remarks bz. ~o Frie~ and lire Mille,!:: Re-statement ot Matter at Issue~ Following a. b1•ie1' summary of the discussion by Hr., Friedman, lira Miller stated tha.t the matter at issue concerned the remark he had made that "a vast amo1mt ot intelligence had been obtained from. the German reading ot the American l!.-.209 JP.s.chine 1•,, and it was now a question of "Whether in tne light or tJse discussion he should withdraw his et..atem.ent in whole or in part or ''hether he would de.fend ito lire Miller considered t.ha.t he should withdraw the P.ord nvastn and asked that the tollolfing statement be substituted.: "Disregarding informa.tion obtained .trom the use of captured ~209 k.a;y lists,, it is generallJ' agreed that a considerable amount of American Jil=.209 traffic was broken by the Germana during World War II, from which little tactical intelligence \1as deduced, but from which an appreciable amount ot strategic intelligence waa obtainedo It ia .further agreed tha.t, due to the inetticiency of the German intercept organization, the lack of modern eryptanazytic machinery. and the failure ot the GerJl'o.ans at high levels to appreciete the value or speedy intercept, for~arding, and processing ot traffic,, the German success did not represent the max.in.um. potential exploitationo 11 General agreement: The above otatement was ;...greed upon by all present a& representing a tair and reasoned resolution or the matter at iasueo Potential Usefulness ot the Discussion; It v1as agreed that a rec.Ord ot the discussion should be made, and that such a reco1•d might be usetul tor: (a) Future Briti5h=American discusaione as ~o the use of the CCll at a combined Army echalo11 lower th.an that at which it 1.s presently used; (b) Background as to the possibl.e solution by potential eneiey agencies or systems such as the Jl...209 and comparable British systems; (c) A source ot reference for discussions pertaining to the proposed. improved Hagelin device and its possible solution by cryptanalytic means; (d) A record ot inte1 est in it£>elf' which should be added to the 1 pertinent operational and historical records or the participating Agenciese Approved: T. 1i.. w. Burton Jliller