The Venona project

Collecting the cables was the easy part. Soviet intelligence stations in the United States sent their important messages home over commercial telegraph lines. As war approached in 1939 the US Army began collecting the enciphered telegrams, and soon thousands were piling up at its Signal Security Agency^1,4. Reading them was supposed to be impossible: the Soviets encoded each message from a codebook and then enciphered the code groups with a one-time pad, a system that is unbreakable if every pad page is used once¹.

A small program at Arlington Hall

On 1 February 1943, on orders from Colonel Carter Clarke of the Army's Special Branch, the agency set up a small unit at Arlington Hall in northern Virginia to study the Soviet traffic. Clarke's concern was that Moscow might negotiate a separate peace with Berlin¹. Gene Grabeel, a former Virginia schoolteacher, started the work and stayed on it for 36 years; she and her colleagues spent months sorting the telegrams by circuit until five separate cryptographic systems emerged¹.

The break came from a manufacturing error in Moscow. For a few months in early 1942 the KGB's cryptographic center printed duplicate copies of the key on more than 35,000 pages of one-time pads and bound them into separate pad books¹. In autumn 1943 Lieutenant Richard Hallock, working on the trade-mission traffic, found evidence of those duplicates. In November 1944 analysts solved the problem of finding where each message started in its pad, and by matching pages used twice they recovered the additive key for parts of hundreds of telegrams¹. Cecil Phillips of NSA later wrote that the breakthrough was "a purely analytic accomplishment, achieved without the benefit of either Soviet codebooks or plain-text copies of original messages"¹.

Reading the KGB

The key alone was not enough; analysts still had to rebuild the codebook. The linguist Meredith Gardner, who had worked on German and Japanese systems, broke the KGB codebook's "spell table" for English words late in 1946. Almost at once he was reading a December 1944 message that listed the leading scientists of the Manhattan Project¹. By May 1947 he could see that the cables described Soviet espionage on a large scale¹.

In late summer 1948 the Army quietly brought in the FBI. Special Agent Robert Lamphere became the link, swapping leads and identifications with Gardner¹. Britain joined from 1948, and the material was given the control term BRIDE in 1950, DRUG in the late 1950s and VENONA only in 1961¹. The National Security Agency, formed in November 1952, inherited the work; the CIA received its first translations in 1953¹.

Names behind the covernames

Venona's yield came mostly from messages of 1944 and 1945. The covernames were then matched to real people by investigation⁴:

• Donald Maclean. From February 1949 analysts saw British Foreign Office telegrams quoted in KGB traffic; by May 1951 the source HOMER had been narrowed to Maclean, who fled to Moscow with Guy Burgess¹. See The Cambridge Five.
• Klaus Fuchs. In September 1949 the FBI matched REST and CHARLES to Fuchs, author of a paper quoted in one message¹.
• The Rosenberg network. In June 1950 the FBI found that LIBERAL, earlier ANTENNA, fitted Julius Rosenberg^1,5. A cable of 21 September 1944 names Ruth Greenglass in clear text, and one of 27 November 1944 describes LIBERAL's wife, "first name ETHEL, 29 years old"².
• Theodore Hall. A cable of 12 November 1944 records the recruitment of Theodore Hall by name. The FBI questioned him in 1951, but he was never prosecuted, probably because the case would have exposed the codebreaking^1,2.

The decrypts also corroborated much of what Elizabeth Bentley and Whittaker Chambers had told the FBI. Judith Coplon, a Justice Department analyst, was the first person arrested on a Venona lead, in March 1949; at her trial the government concealed the source and called it a "confidential informant"¹.

The secret that leaked

The Soviets knew about Venona long before the American public did. William Weisband, a Russian-speaking "linguist adviser" in Arlington Hall's Russian section, watched Gardner extract the list of atomic scientists; the CIA and NSA editors believe he told Moscow about the work. Weisband always denied spying and was never prosecuted for it, though he was sentenced to a year in prison for contempt after skipping a grand jury hearing¹. From autumn 1949 Kim Philby, the British intelligence liaison in Washington, received actual translations¹. By then most of the duplicated pages had been used, so Moscow could do little except warn exposed agents¹.

Inside Washington the secret was kept so tightly that the decrypts were never used in court, and Benson and Warner found no definitive evidence that President Truman was ever briefed on the program¹. The editors argue that this secrecy carried a price: public debate over Soviet espionage in the 1950s was polarized because the evidence the government held could not be shown, a judgment the Moynihan Commission repeated in 1997^1,5.

Closing and release

New messages became rare after 1948, but the work went on, with translations still being issued in the 1960s and 1970s. The program was formally closed on 1 October 1980^1,4. In July 1995, at a ceremony at CIA Headquarters, Director of Central Intelligence John Deutch released the first group of translations: 49 messages on atomic espionage. Five more releases followed, and by 1996 some 2,900 messages were public^1,3.

Read next

September 1941 – June 1950 Soviet atomic espionage (Enormoz) How Soviet intelligence penetrated the Manhattan Project under the codename ENORMOZ, who its sources were, how Venona and confessions exposed them, and what the stolen secrets were worth. 5 September 1945 – 27 June 1946 The Gouzenko affair In September 1945 a GRU cipher clerk walked out of the Soviet embassy in Ottawa with over a hundred documents, exposing a Soviet spy network in Canada and a British atomic scientist. OPERATION · 1934 1934 – 23 January 1963 The Cambridge Five Five Cambridge graduates recruited by Soviet intelligence in the 1930s reached MI6, MI5, the Foreign Office and Bletchley Park, and were exposed one by one between 1951 and 1990. 27 November 1944 A Venona decrypt: New York 1657, 27 November 1944 A short KGB cable from New York to Moscow describing LIBERAL's wife, "first name ETHEL." Broken by Army codebreakers years later, it became the most argued-over Venona message in the Rosenberg case. 1917–1980s One-time pads The one cipher proven unbreakable if its key is random and never reused. Soviet intelligence relied on it, and a wartime printing shortcut that duplicated pages let US codebreakers read thousands of messages. 1914–1990s Signals intelligence Intelligence from intercepted communications and electronic emissions, from Room 40 and the Zimmermann telegram to Ultra, Venona, the Berlin tunnel and NSA's domestic programs exposed by the Church Committee.