Tordella wanted desperately to develop a Radiations Operations Committee of his own with NSA in control, and in October 1961 he invited Hugh Alexander, Hugh Denham, Ray Frawley, and me to Washington, along with Christopher Phillpotts, the MI6 Station Chief, for a special conference to discuss the British cipher breakthroughs. He also invited the CIA and FBI in the hope that by listening to our descriptions of the work of ROC, they would appreciate the benefits of closer cooperation.
I realized from the start that this conference was a priceless opportunity for Britain's Secret Services to redeem themselves in the eyes of the entire American intelligence establishment. The CIA was, by 1961, the dominant intelligence voice in Washington, and although powerful figures there viewed the Anglo-American intelligence alliance as a sentimental luxury in an increasingly unsentimental Cold War, I was confident that if we could demonstrate to them at working level the technical advances made since 1956, we would convince them we were worth cultivating.
Hugh Alexander knew, as I did, that we were taking a gamble. There were no guarantees that the Americans would tell us anything in return at the conference; indeed, it was likely they would not. There were obvious security considerations, too. But the potential gains were enormous. At the very least, we could remove the shadow cast over Anglo-American intelligence relations since the Philby/Burgess/Maclean affair. More important than that, Hugh Alexander had plans for developing the cipher-breaking side of ROC, and I for developing the counterespionage side, which would be possible only with the resources and backing of the Americans. As with the development of the atomic bomb in World War II, we needed to persuade the Americans to fund our ideas into reality. In the long run we would gain the benefits, as the intelligence would flow back to us through the GCHQ/NSA exchange agreement.
The conference was held in specially swept rooms inside NSA headquarters at Fort Meade, Maryland - a vast glasshouse surrounded by electric wire fences, and topped with the tangled arthritic stems of hundreds of aerials and receiver dishes, linking it to the hundreds of NSA listening posts dotted around the world. Louis Tordella and his top cryptanalyst, Art Levinson, attended for NSA. The FBI sent Dick Millen and Lish Whitman; the CIA was represented by Jim Angleton and a bull-like man named Bill Harvey, who had recently returned to Washington to run Staff D, after running the Berlin Tunnel operation.
Harvey was already a living legend in the CIA for his hard drinking and his cowboy manners. He began his career handling Soviet counterespionage for the FBI, until Hoover sacked him for drunkenness. He promptly took his invaluable FBI knowledge and put it to work for the fledgling CIA, becoming along with Angleton one of the most influential American operators in the secret war against the KGB. Through most of the 1950s he served in Berlin, running agents, digging tunnels, and taking the battle to the Soviets wherever possible. For him, the Cold War was as real as if it had been hand-to-hand combat. But for all his crude aggression, Harvey was smart, with a nose for a spy. It was he who first fingered Philby in the USA after the defection of Burgess and Maclean. Harvey had amazing recall for the details of defections and cases decades before, and it was he, before anyone else, who put together the contradictory strands of the MI6 man's career. While others paused for doubt, Harvey pursued Philby with implacable vengeance, and the incident left him with a streak of vindictive anti-British sentiment.
The five-day conference began inauspiciously. Tordella was anxious for a free exchange of ideas, and discussed one or two laboratory experiments NSA was conducting into possible ways of breaking embassy ciphers in Washington, remarking pointedly that in view of the FBI charter, they were unable to go beyond the experimental stage. The CIA and FBI boys were uneasily silent, neither wanting to discuss technical developments in front of the other, or the NSA, or in the CIA's case, in front of us. Angleton took copious notes, while Harvey slumped in his chair with ill-concealed hostility, occasionally snoring loudly, particularly after lunch.
"The Company [the CIA] is here in a listening capacity only," he snapped on the first morning, "We do not discuss our secrets in open session!"
Things began to improve when I read a long paper describing the success of ENGULF against the Egyptians, and the advances we had made since then on the radio illumination of cipher noises and what could be achieved aurally, using our new range of microphones. I went on to give the details of STOCKADE, and at last discussion began to flow. Even Harvey shifted in his seat and began to listen.
On the third day Richard Helms, then CIA Director of Plans, took the chair for a discussion about ways in which these new techniques could be applied to Russian ciphers. I argued strongly that we had to predict the next generation of cipher machines the Russians would develop, and begin work immediately trying to crack them. The non-scientists present were skeptical, but I pointed out that we had done just this during the war at the Admiralty Research Laboratory, when we predicted the new generation of German torpedoes and mines, and were able to counter them as soon as they came into operation. By the end of the discussion, NSA and GCHQ had committed themselves to begin work against the new Russian "Albatross" class cipher machine.
Hugh Alexander was much more interested in the implications for cryptanalysis of the new generation of computers being developed in America. He was obsessed by the Ergonomic Theory, which held that the production of truly random numbers, even electronically as in a cipher machine, was a mathematical impossibility. Alexander believed that if sufficiently powerful computers could be developed, no code, no matter how well enciphered, would be safe, and for the next decade a vast joint research program began to investigate the whole area. (According to a 1986 report in the GUARDIAN newspaper, advances in Ergonomic Theory since 1980 have revolutionized cryptanalysis in the way Alexander predicted.)
As expected, the CIA told us next to nothing about the state of their technical intelligence. They gave the impression that we were not to be trusted with their secrets, but we suspected there were probably other reasons for their reticence. Harvey's Staff D was almost certainly a department designed to bypass the terms of the UK/USA agreement, which specified the total exchange of SIGINT intelligence between NSA and GCHQ. If the Americans wanted to mount a cipher attack and did not wish to share the product with us, or if they wanted to operate against the UK, or a Commonwealth country, as we were sure they were doing, Staff D was the obvious place from which to do it.
Nevertheless, the conference was a milestone in Anglo-American intelligence relations. For the first time in a decade, all six intelligence services sat down and discussed at length how they could cooperate on a wide variety of problems. Major joint research programs were launched, particularly in the computer field, and we had taken a first step in breaking down the walls of mistrust.