CHAPTER 9 “We Just Come Up with the Hardware”
“We Just Come Up with the Hardware”
On the evening of May 27, 2011, a soccer match was about to be held in Moscow between Anzhi, a visiting team from Dagestan, and Lokomotiv, a popular Moscow team. The game was set for early evening so fans could come after work to the modern Lokomotiv home stadium that accommodates twenty-eight thousand in the east of the city. The visiting team, Anzhi, was generously funded by a wealthy oligarch and enjoyed crowds of enthusiastic fans. More than fifteen hundred of them arrived at Lokomotiv stadium that evening from Dagestan to see their team play. They were almost entirely men in their twenties and thirties who wanted to sit together, and they were required to use a separate entrance with high security. They arrived at the gate, holding tickets, and approached what looked to be a rather unremarkable metal detector and some city police. The visitors were from the North Caucasus, where Russia had fought two wars against rebels in Chechnya, so their arrival in Moscow was met with a apprehension and scrutiny. The police routinely patted them down, looking for weapons.
The fans coming through paid little attention to a camera sitting on a tripod aimed at each of them as they stepped up to the metal detector. The lens was aimed at their faces. The camera rapidly attempted to capture each face into a green digital frame and then identified different characteristics of the face, including such distinctive features as distance between the eyes. Then the camera went to work, snapping several photographs of each face. A computer connected to the camera then evaluated each person based on a complex algorithm, and within seconds the person’s name was established and they were given a unique number. They may have all come to see a soccer match, but they had also just walked into a modern and potentially powerful system of face recognition. Originally invented to help spot criminals, face recognition had expanded in the hands of security services to be a tool for surveillance of all kinds of people at any kind of public event and in public places.
Near the metal detectors sat an operator with a laptop who worked for a company called Ladakom-Service, and he monitored every face closely. One window on his screen showed the live camera acquiring the face images, another part of the screen showed the captured images, and a program was constantly running to match the captured images with people in a government passport database, one of the biggest in the country. When the match was successful, a photograph just taken appeared along the bottom of the screen with the person’s full identity. The government had obtained a current picture and identification of thousands of people that could be used for almost any purpose in the future.
It happened not only at sports events. The same company in 2011 had installed this technology in the entrance hall of one of the busiest metro stations in the city. The station, Okhotny Ryad, is located a stone’s throw from the Kremlin and around the corner from the Russian lower house of parliament, next to one of the most heavily traveled streets in Moscow. As people stepped on the subway escalator, their faces entered a frame and were captured by video cameras. The images were rapidly linked to their identity in security service databases. There was no notification to anyone that they were being recorded.
The system was so advanced that a scan of 10 million images would take no more than seven seconds. The facial images and video are sent to the Metro system’s situation room, the Interior and Emergencies Ministries, and to the FSB.
The facial recognition system is a glimpse into a large and mostly hidden phenomenon that was a profound legacy of the Soviet experience: the use of engineering to build systems for the security services to control information and populations. These systems were invented and developed by engineers who knew what the systems could do but rarely if ever questioned the purpose of control for which they were used.
The Soviet Communist Party held a monopoly on power and did not want competition. It imposed rigid conditions on all kinds of people; for engineers, there was pressure to conform to the goals of the party-state and to fulfill its technical needs. To succeed meant to work on projects without questioning the big picture. For many decades Soviet engineers were schooled intensively in technical subjects but rarely if ever had exposure to the humanities; the breadth of their education was exceedingly narrow. Unlike medical doctors who were trained in ethics, engineers were not. They were taught to be technical servants of the state. As a result, generations of engineers were trained and worked their entire lives with little understanding of politics or trust of politicians and were suspicious of public activity as a whole. These engineers were focused on the immense technical needs of the Soviet Union and were comfortable with the concept of strict order because it suited their understanding of the mechanical world better than the often-unruly reality of freedom.
When Stalin’s security services in the 1930s and 1940s needed to conduct secret research in particular areas, they arrested scientists and engineers and sent them to special installations, the sharashkas, which were closed off from the outside and heavily guarded. The scientists and engineers were motivated to produce quick results under the threat of being sent to labor camps if they failed. But in the years after Stalin’s death in 1953 this system evolved into a far-reaching system of research institutes, not all of them closed. A result was that many thousands of Soviet engineers were working in security or military research.
The sharashka in Marfino, east of Moscow, was especially important, and by 1948–1949 it had become a relatively large research effort, with 490 personnel, of which 280 were prisoners, divided into twelve research groups, including the acoustic laboratory of Major Abram Trakhtman.
Marfino’s main task was to develop a special kind of secure telephone system that would allow Stalin to speak on the phone without interception. To accomplish that, the voice on the phone would have to be split into pieces, coded, and then reassembled. The problem was not only how to code the pieces but also how to repair the speech in a way so that the speech and speaker would be instantly recognizable. For months Marfino’s acoustic laboratory tried to make the decoded speech recognizable. The effort required the focus of both inmates and their superiors. They worked under the guidance of Trakhtman, but the real brain behind the research effort was Lev Kopelev, an inmate. He was considered the top expert on recognizing speech, with an unerring ear for accents and a deep grasp of the physics of sound waves.
In the late autumn of 1949, at the yard of Marfino, Kopelev approached his closest friend, Alexander Solzhenitsyn, a fellow inmate, to share with him a state secret. Kopelev had successfully listened to and identified a caller who gave sensitive information to the US Embassy about the atomic bomb. Kopelev was incensed.
Although he had been convicted and sentenced for expressing his disapproval of the Soviet troops’ harsh treatment of the German population in 1945, Kopelev remained a devoted Communist and Soviet patriot, and he was outraged to listen to someone who had just betrayed such a big secret to the Americans. He was given four audio tapes—it turned out the speaker tried three times to reach the US Embassy and then gave up and called the Canadian Embassy once. The security services intercepted and recorded all the conversations. Kopelev was also given samples of phone conversations of three suspects. With his considerable skills and talents, Kopelev pointed the finger at one of the recorded voices, leading the authorities to a certain foreign ministry official, who was arrested. It was a major—and unexpected—victory for the sharashka. Excited, Kopelev couldn’t help but tell his friend Solzhenitsyn what he had done. Kopelev coined the term phonoscopy for the new scientific discipline of recognizing the identity of a speaker on the phone, one that would be very valuable to the security services for many years to come.
In July 1950 the secure telephone technology for Stalin was finished, as was the main assignment for Marfino. For the next two years Marfino created a manufacturing line for the equipment they had invented. When that was completed, the sharashka was essentially divided in half. The specialists on secure telephony were left there, and it was renamed to become the top-secret National Research Institute No. 2, working on the protection of Kremlin telecommunications, as it does to this day. The other half of specialists, who had worked in the acoustic laboratory, including Kopelev, were moved to Kuchino, another sharashka located outside of Moscow.
This was a fateful move that established the KGB’s central role in research and development of listening devices and eavesdropping for the next half century. The transfer to Kuchino meant that the secret services would do the research on speech recognition technology in the same facility—and guided by the same people—as those working on wiretapping. They wanted to make sure they could not only intercept a conversation but also have the means to identify those who spoke on the phone. The KGB wanted full control of telecommunications, and from this time on, identifying a speaker was considered a legitimate part of Soviet surveillance.
Kuchino, surrounded by high walls seventeen miles east of Moscow, was the main research facility for Stalin’s secret services in the area of special, or “operative,” equipment—ranging from weapons to radio sets to, most importantly, listening devices. In one of their most ambitious and successful exploits, the experts at Kuchino planted a listening device inside a large replica of the Great Seal of the United States and presented it as a gift to the US ambassador in August 1945, and it was hung in the ambassador’s study. The device transmitted sound waves out of the ambassador’s study to the Soviet secret services until it was exposed in 1952.
Kopelev left Kuchino in 1954, a year after Stalin died, never to return to his research in phonoscopy. He went on to become a dissident. But he left his archives behind in Kuchino, and they were carefully preserved. For a time the security services didn’t know what to do with them; it seemed that the technology Kopelev invented was based on his knowledge alone and wouldn’t work without him. In other words, without Kopelev’s unique skills, it was useless to try to identify a speaker on the phone.
But other research began to show that there was a method behind Kopelev’s success. The first evidence came in 1960 when a Swedish scientist, Gunnar Fant, published a monograph, The Acoustic Theory of Speech Production, based on his research at MIT. He had found a way to slice up a voice recording into samples and then identify them using mathematics and physics. This meant that there was a more reliable and verifiable scientific method instead of relying on Kopelev’s skills. Fant’s discovery, translated into Russian in 1964, led to a surge of secret research into the topic inside the Soviet Union.
While the scientists pursued his theory, Fant began to be concerned that law enforcement would abuse speech recognition technology. Fant’s concerns were confirmed in 1970, when FBI director J. Edgar Hoover went to Stockholm and gave an interview to the local newspaper Dagens Nyheter about shining prospects for using voice samples for identifying terrorists. When the newspaper asked Fant for comments on Hoover’s remarks, Fant cautioned that the method he had developed was imprecise and it was premature to use to identify anyone. His rebuttal was so surprising that the newspaper printed on the front page photographs of Hoover and Fant opposite each other, presenting him, as Fant put it later, as a “possible FBI enemy number one.”
Soviet scientists had no such reservations. Research centers working on speech recognition opened in many cities, and the section on acoustics at the Academy of Sciences coordinated the nationwide research. But everybody knew that the true boss was the KGB.
An instrumental part of the research was in Leningrad, the Scientific Research Institute of Dalny Svyazi, or of long-distance communications, known as Dalsvyaz. This is the facility where Sergei Koval, a graduate of the physics department of the University of Leningrad, began work in 1973 on acoustics. He was always interested in the science of sound, but what was also attractive was a promised monthly salary bonus of 15 percent. He was unconcerned that the institute was shrouded in secrecy. The offices of his applied acoustics unit were always guarded by men with automatic weapons. The institute, with more than ten thousand personnel, was overseen by a ministry for industrial telecommunications, but its real purpose was to work for the military. The applied acoustics unit of three hundred people that Koval joined was not under control of the institute at all but was instead run by the KGB, who paid these additional personnel the bonus. It was a classic Russian matryoshka—secrets within secrets—applied to research.
Koval soon realized the reason for such secrecy. His colleagues told him that this unit was in fact the Marfino sharashka that had been transferred to Leningrad. One day he was pointed to a bespectacled engineer who worked at a neighboring laboratory. His name was Valentin Martynov, and he had once served in Marfino along with Kopelev and Solzhenitsyn (and featured in Solzhenitsyn’s novel The First Circle under the name “Walentulya” Pryanchikov). Koval recalled that Martynov was “meticulous and stubborn.” A young, enthusiastic engineer of the late 1940s, Martynov had remained devoted to speech recognition for decades. He went so far as to defend a thesis for a degree on the topic. Although he was free now, he still walked every day into the territory guarded by the men with automatic weapons and dogs to do research for the same secret services that had once sent him to prison. Koval never tried to ask him why: “It was a generation that was much more mature. It was not suitable to talk about the past.”
By the 1970s Koval’s applied acoustic unit became the main coordinator of research funded by the KGB in speech recognition. He recalled, “There was a section of applied problem-solving at the Academy of Sciences. This section took orders and research commissions on prospective research from all the agencies, from the Ministry of Defense and from the KGB. The section demanded money, and it always got the money. The scheme was wonderful: the money then was allocated to the applied research departments belonging to the KGB, like our department. So we were able to distribute this money right across the different academic institutions as we saw fit. We could effectively sponsor any project we wanted. I myself was the curator of the scientific program, where forty universities were involved.”
What began in the 1940s with seven people in the acoustic laboratory in Marfino had, by the 1970s, become a sprawling, well-funded empire of secret research. There was no clear line between KGB-sponsored research and civilian research; it was all part of the same empire. The secrecy touched everyone—numbering millions of people. It showed up in the most unexpected places. For example, the Computation Center of the Academy of Sciences on Vavilova Street in Moscow, which Ed Fredkin loved to visit in the 1980s to talk personal computers, was one of the research institutes quietly working for Koval’s unit.
Vladimir Chuchupal joined the section of voice recognition of the Computation Center in 1980. He was told that the main task of the section was to apply computers to speech recognition. Chuchupal was warned that it was strictly prohibited to mention to anyone outside the Center the name of their main “customer”—the Dalsvyaz and the KGB. He was put in direct contact with Kuchino almost immediately. Chuchupal knew exactly what they worked on—one day his chief described how he was given notes from the legendary Kopelev to study.
Thanks to the generous funding provided by the KGB, in the early 1980s Chuchupal’s section got its first personal computers, some Soviet-made machines and a few IBM PS2s. When they arrived, the issue of speech recognition opened up a vista for surveillance the KGB had never imagined possible—applying computer technologies to phone tapping meant that not only could a speaker be identified but that what he said could be used to trigger the interception system (the surprising byproduct of the project was the computer game Tetris, designed on one of the KGB computers). That, at least, was the theory. The KGB came up with the idea of using key words so that mention of “the bomb” or the “Communist Party,” or anything else chosen by the KGB and put in the system would automatically initiate interception of the phone line. This option could have changed the KGB’s modus operandi completely—in most previous cases the KGB needed to know the identity of the suspect to start eavesdropping on his phone; now the technology would provide the suspects. But it was also very challenging; the keyword system was an ambition, but making it a reality depended on more computing power than was available.
The Soviet research empire into speech recognition along with Dalsvyaz and the unit at the Computation Center worked actively on the issue for years. Once again, there was no clear line between the civilian and KGB research.
This empire cracked with the collapse of the Soviet Union, but it didn’t go away. Initially the KGB cut its research programs. “In 1990 our funding stopped,” Koval recalled, who himself left Dalsvyaz. “Two-thirds of employees quit immediately.” But he did not leave the field of research. With his laboratory chief, Mikhail Khitrov, and five colleagues, he founded a private company that in 1993 became the Speech Technology Center, trademarked in the United States as SpeechPro. Each of eight founders got an equal share of 12.5 percent of the company. The speech recognition scientists tried to succeed with civilian contracts; for example, they developed a talking book for the Society for the Blind.
But soon their old friends, the security services, returned. Koval’s company got its first contract from the Interior Ministry to build a system of using phonoscopy for chasing criminals. Then the FSB offered a contract to make a system that would separate voice from background noise. In the 2000s the company employed up to 350 people—roughly the size of the original Soviet department in Dalsvyaz. “I cannot say what kind of work we do for them,” Koval said of the security services, “but it all continues, it’s the same—what we did then, we do now.”
The company has developed technology they consider unique in its capability and reach. It is able, for example, to store many millions of items of biometric data, such as voice samples and photo images, and match them to individuals by searching the world’s communication channels, including video files. The voice recognition technology can identify the speaker, regardless of language, accent, or dialect, based on physical characteristics of the voice.
In 2008 the company completed its first national voice recognition project in Mexico. The system was able to use state records of human voices and biometric details—voice, face, and other characteristics—to identify individuals, and to do it from fragments of speech alone. Mexico’s national database of voices was made up of speech fragments recorded from criminals, law enforcers, and many law-abiding citizens, who are obliged to supply vocal samples for state regulated activities, such as obtaining a driver’s license. Thus, Kopelev’s 1949 dream of creating the system that “would allow recognition of the voice in all circumstances out of any amount of voices” was realized in 2010 in Mexico. Koval was personally in charge of implementing the ambitious project. “I’ve been traveling to Mexico for seven years!” he exclaimed.
On a cold and snowy day in January 2012, in an almost empty caf? near Chernyshevsky Metro Station in St. Petersburg, Koval enthusiastically recalled to Andrei the story of his company. Koval proudly listed countries where his company’s speech-recognition technology was already in use: Kazakhstan, Kyrgyzstan, Uzbekistan, and Belarus—all repressive authoritarian regimes in the former Soviet Union—as well as Saudi Arabia, Algeria, Yemen, and Turkey.
Andrei asked Koval what he thought about the ethics of his work and the fact that regimes around the world were using his technology to suppress dissidents. He replied emotionally and with certitude and self-confidence. “All this talk about technology catching dissidents is just bullshit,” he insisted. “It’s typical of the kind of psychological warfare the Americans use against their opponents. I think all these arguments about human rights are completely hypocritical.” He expressed no reservations about the use of his technology against journalists, dissidents, and human rights campaigners. “What can we do about it?” he said. “We just come up with the hardware. It’s just technology that is developed with law enforcement in mind. Sure, you can use it against the good guys just as easily as you can use it against the bad guys. One way or another, these governments will be able to use surveillance technology, whether we supply it or not. Take, for example, face-recognition technology: you can film a demonstration, and with that film you can identify the journalists, the drug addicts, the recently released prisoners, or the nationalists. It’s all the same technology. I can’t think what can be done about that! If governments listen in on people’s conversations, it’s not the microphone’s fault!”
These exact words have been repeated over and over again by engineers who willingly served the Soviet state and then did the same thing in Russia. They believed it was not their fault. Koval’s confidence had recently been bolstered by an investment from a source even closer to Putin than the secret services could provide. In September 2011 Gazprombank acquired 35 percent of SpeechPro. Gazprombank is also part of the vast business empire of Yuri Kovalchuk, a close friend of Vladimir Putin.
The Russian system of secret research appears to be reestablished completely. In Moscow, Chuchupal, now the chief of the sector on speech recognition at the Computation Center of the Academy of Sciences, continues to work on speech recognition, and Kuchino is among his customers. Both Koval’s and Chuchupal’s organizations are still working on the issue of “key words.”
Koval’s odyssey was repeated over and over again by other Soviet scientists and engineers, and it created a mindset among many of them. Loren Graham, a preeminent historian of Soviet and Russian science at MIT, told us, “Russian scientists and engineers are, on the whole, less interested in the ethical and moral problems of their work than many of their counterparts in Western countries.”
“Why is this so? I see two reasons,” he added. “In the Soviet period Russian scientists and engineers learned early on that if they raised ethical and moral issues that this was seen by the authorities as ‘political opposition,’ and they would be punished for raising such issues. Therefore, they learned to stay silent, and after a while this silence became ingrained and even a part of their professional definition. Of course, the Soviet Union is long gone, but these attitudes have largely continued.”
Second, he said, “Engineering education in Russia has been focused on technical issues, with very little attention to larger human, ethical, and moral questions. Although engineering education in the US has some of these characteristics also, it is worth noticing that at top engineering schools in the United States, such as the Massachusetts Institute of Technology—my university—every student during four years of engineering study is required to take eight courses, usually one each semester, in the humanities and social sciences. These courses open up deep questions of ethics and ‘meaning,’ which are not considered in technical courses.
“This is an important part of engineering education in the best universities in Western countries,” Graham said. “It has important effects, leading to questions about the social responsibilities of scientists and engineers. And many of the best engineering schools in the United States also have departments of Science, Technology and Society [STS], where these problems are studied.”
Anatoly Levenchuk, an engineer himself who, in the early 1990s, helped launch Relcom, told us that “I tell my students not to apply system engineering when you work for the government.” Why? “It could be very dangerous. You need to know humanities to deal with the state. If you apply only engineering, you will build a prison as a result. Say you are tasked to address threats, in this case the best way to address them as engineer is to build a box, a prison, you just close everything off.”
Levenchuk himself ceased to cooperate with the government in 2006 and focused on teaching engineering at Moscow Institute of Physics and Technology, the major university for training engineers for the Russian nuclear industry. Levenchuk sought to present new ideas in his classroom, challenging the traditional rigid approach and urging students to be more open-minded and aware. He was soon attacked. In April 2013 he was confronted with an old and nefarious Soviet practice—a public denunciation. An open letter was published on LiveJournal.com, accusing Levenchuk of teaching “fascist” philosophy and values. Levenchuk was accused of systematic destruction of “the Soviet school of design.” The denunciation then went on to demand that members of the State Duma to initiate a request to the General Prosecutor’s Office to check whether Levenchuk is a foreign agent, as “pro-Western ideas” are detected in his lectures. The request was duly sent, and the institute was forced to write an official reply, protecting Levenchuk.
The denunciation showed that Soviet engineers’ mindset—the rigid adherence to the technical—was resurging under Putin.
Sergei Koval, so dismissive of concerns about dissidents and human rights, took his approach overseas in 2009. It was one thing to loyally and unquestioningly serve the state in the former Soviet Union, but it was quite another to deliver the same approach to other countries and security services. Koval’s journey abroad took him to Colombia. There, on September 21, 2009, the secret police, Departamento Administrativo de Seguridad, or DAS, a hybrid of intelligence and law enforcement, held a press conference in Bogot?.
For a year DAS had been under constant criticism, accused of illegal wiretapping of journalists, opposition politicians, human rights groups, and even Supreme Court justices. The scale of the scandal caused journalists to coin the term “Colombian Watergate.”
In the announcement of the press conference DAS promised to present some new crucial evidence to address the question of illegal wiretapping. The DAS director, Felipe Mu?oz, a thirty-nine-year-old energetic technocrat, trained at the London School of Economics and Colombia University, appeared at the press conference with Koval sitting alongside.
Mu?oz announced that DAS had conducted the internal investigation and invited an independent expert from Russia, with more than thirty years experience in the field, to examine leaked recordings of intercepted calls and compare them with the recordings made by DAS legally. He tried to prove that the agency was not involved in illegal wiretapping. Then he presented Sergei Koval.
Koval stated that he had applied more than twenty various audio characteristics in the course of examination, and the analysis showed that the recordings had been made using completely different types of equipment. “The conclusion was unequivocal: these wiretaps were not registered with this type of equipment,” Koval added.
If true, this would absolve the security service from having made illegal wiretaps. Koval claimed the Colombian secret service didn’t have the necessary equipment to produce the type of recordings leaked to the media. Mu?oz, in turn, was happy to point a finger to some private unidentified spies, “We have sufficiently strong preliminary evidence to say that there is a market for mobile equipment interception which lacks control.”
Koval had come halfway around the world to speak up for the Colombian secret police. And it turned out these public declarations were wrong. A few months later a prosecutor in Colombia declared that he had proof that a DAS team had spied on public figures with the knowledge of officials in President ?lvaro Uribe’s office. Eventually some DAS officials confirmed that it was indeed the DAS that had conducted illegal wiretapping. One employee admitted he had received orders directly from the DAS director, Maria del Pilar Hurtado, and the intended recipient of wiretap transcripts was President ?lvaro Uribe. In late 2011 the agency was finally disbanded, and the expert on speech recognition was long gone from Bogot?, back to Russia, and then on to Mexico.
On the outskirts of St. Petersburg, in a glossy new business center, there is a small company named Protei. The company’s office was a bit chaotic in 2011—they had just moved in—with tables and wires all over the place. The computers had yet to be installed, but Protei was already making something highly desired outside of Russia: the equipment for making sure that the black boxes—the SORM technologies—would work in countries like Uzbekistan and Kazakhstan, where authoritarian rulers with miserable human rights practices and intolerant of democracy and dissent were eager to use the technology to control the Internet. The company produced all kinds of technology from SORM-1 to SORM-3, from phone eavesdropping to Internet intercepts. In December 2011 WikiLeaks and Privacy International launched the Spy Files project, a database on companies that sell such surveillance gear around the world. Although most of the vendors are British, Israeli, German, and American companies, it also included Koval’s SpeechPro and Protei.
We went to see another engineer who had made it in the world of secret services and secret surveillance. Vadim Sekeresh was head of the SORM department at Protei. A phlegmatic, forty-year-old graduate of the applied mathematics department of St. Petersburg University, he seemed unruffled by the WikiLeaks disclosure. Like so many other engineers, he did not ask deep moral or ethical questions about how his products were being used. “I didn’t pay any attention to it,” Sekeresh said of the report. “I didn’t really look into it because the whole thing doesn’t bother me. After all, we are not producing the listening devices, or bugs. And… we aren’t the only ones producing such tech anyway.” A few months later he told Andrei in an e-mail, “Lots of crimes are solved thanks to technology. It’s obvious that everything could be used to harm, but it’s not related to the producers.”
In other words, it is not the engineers’ fault.
In 2012, the year Internet filtering was introduced in Russia, Protei developed a product based on DPI technology to implement the censorship of Roskomnadzor. In March 2015 Protei announced that the company had successfully deployed an Internet-filtering system based on DPI on the network of Kyrgyzstan’s telecom operator MegaCom, one of the largest in the country. Russian engineers, once again, developed the hardware that brought one of the world’s most intrusive Internet-filtering technologies to Central Asia.
- CHAPTER 9 “We Just Come Up with the Hardware”
- 4.4.4 The Dispatcher
- About the author
- Chapter 5. Preparations
- Chapter 6. Traversing of tables and chains
- Chapter 7. The state machine
- Chapter 8. Saving and restoring large rule-sets
- Chapter 9. How a rule is built
- Chapter 10. Iptables matches
- Chapter 11. Iptables targets and jumps
- Chapter 12. Debugging your scripts
- Chapter 5 Installing and Configuring VirtualCenter 2.0