Majid Shahriyari became an Iranian martyr while he was driving to work on an autumn day in Tehran. As he made his way along Artesh Boulevard, an explosive device ripped through his car. The 45-year-old was a devout man: Iranians would describe him as a Hizbollahi, a person fiercely loyal to the country’s Islamic system and easily identified by his unshaven face and simple clothes. But Shahriyari also stood out for another reason. He was one of Iran’s leading atomic scientists, an expert on nuclear chain reactions.
Iran has long maintained that its atomic programme is aimed at developing peaceful nuclear energy. But much of the outside world believes its true intention is to build a nuclear weapon. Either way, Shahriyari was indispensable.
On November 29 2010, as the scientist and his wife were on their way to Shahid Beheshti University, where they worked as professors, a motorcycle pulled up alongside their car. The riders then attached an object to the driver’s door window, and sped off. A few seconds later, an explosion blew the door off the left side of the car. Moments before the bang, the couple appeared to have an inkling of what was about to unfold. Shahriyari’s wife scrambled out of the car and survived. But her husband had no chance. As the remains of the vehicle smouldered in the road, the scientist’s body lay slumped on the steering wheel. Later, his wife would tell state television that his death was an honour: “The blood of my martyr is sacrificed for the dignity of the nation.”
The assassination of a man of his standing would not usually trigger headlines around the world. But Shahriyari was not the only victim that day. On the same morning, Fereydoon Abbasi-Davani, Shahriyari’s university colleague and a man close to Iran’s elite Revolutionary Guard, survived a similar attack. Eight months previously Massoud Ali Mohammadi, an expert in quantum physics at Tehran University, died when a booby-trapped motorcycle exploded near his house. The murders triggered a mixture of bewilderment and fear across Iran’s scientific community.
Who is killing Iranian scientists? Two and a half months after Shahriyari’s death, the answer is still a mystery. Iran has pointed the finger at its enemies in the US and Israel, claiming the assassinations are part of a broader campaign to derail their nuclear programme. “Shahriyari was the top when it comes to the nuclear programme; in terms of his knowledge, he was irreplaceable,” says a former Iranian official. “His death is truly damaging to it,” he adds. “The gallows will soon be earmarked for the retribution of the blood of Shahriyari,” Mohammad-Reza Naqdi, head of Iran’s Islamic Basij militia told local media. But in the west, another possible explanation has been offered: perhaps the Iranian regime instigated the killings, taking revenge on men whom it had started to regard as suspect and politically dubious?
What is not in doubt, however, is the timing. The murders have taken place in a period when intelligence agencies in Israel, the US and the west have embarked on a high-risk, and high-tech, bid to sabotage Iran’s nuclear programme – whether or not it is aiming for nuclear capability or to build an actual nuclear bomb.
The death of Shahriyari caused shockwaves across Tehran. But little attention has been paid to its impact in another part of the Middle East, 1,500 miles west of the Iranian capital. Around nine months before his death, Shahriyari became Iran’s representative to an unusual international scientific project to promote Middle Eastern co-operation in science. Established eight years ago in Allaan, a small farming area in central Jordan, the Sesame project centres on an old particle accelerator donated by the German government to study atomic structures. The enterprise aims to bring scientists from Israel, the Arab states and Iran together to run experiments. The German scientists who dreamed it up had in mind a Middle East version of Cern, the European institute created after the second world war to help unite a divided continent. Since the project’s inception, Iran has shown a keen interest in Sesame, sending unusually large delegations, say participants from other countries. Both Shahriyari and Ali Mohammadi were participants, the first replacing the second after his death. “The first killing might have just been coincidental,” says a Sesame council member. “The second killing makes the link which these two men had with Sesame more significant.” Could their involvement in Sesame be related to their deaths?
One possibility is that their interactions in Allaan might have exposed the scientists to suspicion that they were complicit in sabotaging Iran’s nuclear programme. In Tehran’s political and diplomatic circles, the killing of Ali Mohammadi was seen as a possible act of revenge by the regime. He was, after all, a known sympathiser with the Green movement, which had flooded the streets of Tehran in June 2009 to demand a re-run of the Iranian election. Ali Mohammadi’s wife voiced suspicions of her own. “Some people tell me that insiders killed my husband,” she told a website linked to Iran’s domestic opposition.
Reuel Marc Gerecht, a former CIA agent in the Middle East now with the Foundation for Defense of Democracies, says that any Iranian involved in Sesame was “certainly considered for a ‘cold pitch’” – an approach by intelligence agencies. “Jordanian intelligence would have reviewed everyone and that information would have been shared with the Americans if not the Israelis,” Gerecht says – adding that Tehran would have known such approaches were likely. “Here you have two men who were players in the Iranian nuclear programme but were able to meet with outsiders in the Sesame project,” adds one western intelligence official. “My guess is that when things started to go wrong with the nuclear programme, the Iranians started to point the finger at them.”
Inside Tehran, Shahriyari was seen as too valuable for a regime hit. While the government had quickly moved on from Ali Mohammadi’s murder, it appeared more nervous about the loss of Shahriyari and the attack on Fereydoon Abbasi-Davani, listed in a 2007 UN resolution as a senior ministry of defence scientist involved in the nuclear and ballistic missiles programme.
In November, Iran’s intelligence ministry came under pressure to explain how its anti-espionage department had failed to track plots against its scientists. Had the participation of the two murdered men in Sesame alerted Israel to their identity? That might have been a first step towards a possible assassination by Mossad, the Israeli secret service, or allied agents in the People’s Mojahedin of Iran, the exiled group that advocates the overthrow of the Iranian regime. It is considered a terrorist organisation by the US, although it maintains that it is no longer involved in illegal violence.
In Israel, many analysts assumed Mossad had a hand in the killings. Its master spy, Meir Dagan, whose retirement was announced on the same day as Shahriyari’s murder, had been hailed in Israel as one of the secret agency’s most effective chiefs. A short, stocky man with a large, balding head and an unassuming demeanour, Dagan took over Mossad in 2002. At that time the agency was still recovering from one of its worst-ever blunders: the failed assassination of a leader of the Palestinian Islamist group Hamas. Dagan was entrusted with a mandate to build an agency “with a knife between his teeth”, making Mossad live up to its reputation for daring, high-risk operations. And he was to focus, above all else, on Iran’s nuclear programme.
Not all went well under Dagan. The assassination of a Hamas operative in Dubai last year was an embarrassment; although the targeted killing succeeded, it left a trail of clues that allowed the Dubai police to accuse Israel. But for some Mossad watchers, the Tehran assassinations were seen as a dramatic comeback for Dagan’s agency. “The message they send by doing this is that you must try to defect, otherwise you’ll die,” argues one western diplomat. “It’s a good way to spread panic.”
Some intelligence experts, however, doubt that Mossad could pull off assassinations in downtown Tehran by itself, whatever its skills. They suggest that if Israel was behind the killings, it would have used Iranian allies. Bob Baer, a former CIA operative in the Middle East, says that Mossad could not possibly have its own operatives inside Iran and would have had to rely on others to do the killing. “You have to exclude that there are Mossad operatives in Iran, or any Iranian who emigrated to Israel and went back,” he says, adding that a more likely candidate, in his judgment, would be the People’s Mojahedin acting on Israel’s behalf.
A few hours after Shahriyari’s car was blown up, President Mahmoud Ahmadi-Nejad held a press conference. Dressed in a dark grey suit, he fielded questions from the presidential complex. Predictably he launched into a tirade against western governments and the “Zionist regime”, accusing them of the Shahriyari murder. Then, towards the end of the conference, he was asked about the “virus” that had mysteriously disrupted some of Iran’s nuclear facilities. He leaned forward and, very casually, as if referring to some unimportant detail, he made a remarkable admission. “Problems were created for a limited number of centrifuges due to the software installed in some of the electronic equipment,” he said, in reference to the key equipment used to enrich uranium. “Our experts discovered them and today they are no longer able to do such a thing.” Ahmadi-Nejad did not name the software. But there was little doubt: the culprit was Stuxnet – a malicious software code first noticed midway through 2010, and a main tool of sabotage in the secret war against Iran.
The drive to sabotage Iran’s nuclear efforts is long-standing. Even before the 2002 revelation of a uranium enrichment facility at Natanz, the US and Israel worried that Iran could use its programme – begun under the Shah and stepped up in the 1990s – to develop the bomb. The anxiety mounted over the years as Tehran accelerated its advances and the stock of low enriched uranium built up.
The international dilemma over how to halt Iran’s nuclear progress has been one of the main, if not the biggest, foreign policy challenges that the US has faced over the past decade. Even the most hawkish figures acknowledge that the military option – bombing Natanz and other facilities – implies immense risks. Given Iran’s ability to retaliate, it could provoke an uncontrollable regional war. In the latter years of the Bush administration, the US made clear its opposition to a strike; a position that has become more entrenched since Barack Obama entered the White House.
Against that backdrop, the US has embarked on a sanctions drive to increase pressure on Tehran, isolating its economy and making it difficult for the country to obtain materials for its nuclear and missile programmes. This was backed up by a series of United Nations sanctions, which began at the end of 2006. US officials say their efforts have been aided by the relatively primitive enrichment technology Tehran relies on – Iranian copies of Pakistani copies of 1970s Dutch centrifuges – and the limitations of Iran’s own industrial base. “The sanctions are working,” Hillary Clinton, US secretary of state, said in January 2011. “Their [Iran’s] programme, from our best estimate, has been slowed down.” What she didn’t say however, was that sanctions are only part of the picture. Measures adopted at the United Nations and elsewhere are complemented by a sabotage campaign.
The sanctions have left Iran reliant on black market middlemen. They in turn are open to western persuasion to sell defective components, further damaging Iran’s nuclear infrastructure. One senior US official makes a rare acknowledgment of the relationship between Iran’s frantic hunt for nuclear components and tampering by Washington and its allies. “When you have to cast a wider net, there are more opportunities for intervention,” he says. For “intervention” read “sabotage”. In a speech last year, Sir John Sawers, head of MI6, also alluded to western intelligence agencies’ role. “We need intelligence-led operations to make it more difficult for countries like Iran to develop nuclear weapons,” he said. “SIS’s [the Secret Intelligence Service] role is to find out what these states are doing … and identify ways to slow down their access to vital materials and technology.”
Examples of this sabotage go back for years. In one well-documented case in 2000, the CIA used a Russian to provide Iran with an erroneous blueprint for a nuclear weapon. Prosecutors in Switzerland are currently considering whether to issue charges against the Tinner family, former middlemen who traded nuclear components, who were recruited by the CIA as part of its bid to sabotage equipment sent to Iran and Libya. Mark Hibbs, an expert on Iran’s programme at the Carnegie Endowment for International Peace, stresses that one attraction of sabotage is that centrifuge failures tend to have a knock-on effect. “It’s a very violent explosion because these things are operating at supersonic speed,” he said. “When they blow up it is referred to by engineers as fratricide. A centrifuge will explode and the shrapnel will cause other centrifuges operating in the vicinity to break down and crash.”
But despite such efforts, the sabotage campaign never made a big breakthrough. A recently published book, Fallout: The True Story of the CIA’s Secret War on Nuclear Trafficking, by Douglas Frantz and Catherine Collins (authors of a well-regarded study of A.Q. Khan, the Pakistani who built an international nuclear proliferation network), says that in 2004, Urenco (the European group which is the world leader in uranium enrichment) tested centrifuges intended for Libya that the US had tampered with. It concluded the devices had not been sabotaged enough – they still worked. Something else was needed. “If the initial sabotage efforts from early 2000 to 2003 had worked [on Iran’s programme], we [the international actors most concerned about Iran’s nuclear programme] wouldn’t have to ratchet up the danger by releasing something like Stuxnet,” says Frantz. “I don’t know the origins of Stuxnet, but it took the sabotage effort to a new and a very risky level.”
Computer experts have long been aware that governments or other actors can carry out cyberattacks over the web. But Stuxnet was of a different order, an aggressive computer worm designed to destroy a physical piece of industrial infrastructure. The Stuxnet worm takes over computers that control major pieces of machinery, giving it access to the industrial controllers at the plant and items of equipment that operate pumps and valves. Once in the system, it can prove near impossible to dislodge. “Even if engineers discover the worm and disconnect their laptops, it is programmed to continue operating,” Ralph Langner, an IT expert in Hamburg, told the FT last year. For much of the last year, international scientists were unclear exactly what Stuxnet had been designed to attack. But slowly the pieces of the puzzle came together. Although it affected thousands of computers across the globe, Stuxnet seemed specifically designed to attack Iran’s nuclear facilities. There is an emerging consensus that it was responsible for the apparent failure in 2009 and 2010 of about 1,000 of the 8,000 centrifuges used to enrich uranium at Natanz.
Like other experts, David Albright, president of the Institute for Science and International Security (Isis) in Washington, is confident that big western intelligence agencies are behind the worm. “The US and Israel have the manpower, the motivation, the resources to do this,” he says, adding that it is hard to believe Germany was not involved because of the technical knowledge needed to manipulate equipment made by companies such as Siemens, which produces control systems of the sort used at Natanz. Unsurprisingly, US officials take care to dissociate themselves from Stuxnet, which has disrupted companies in countries as far apart as Belarus, Indonesia, India, Ecuador and Taiwan. William Lynn, US deputy secretary of defense, says: “It’s hard to figure out where all these things are coming from.” But such statements fall short of categorical denials.
Hibbs, the expert at the Carnegie Endowment, points out that, in 2003, the US took possession of centrifuges supplied to Libya by A.Q. Khan’s nuclear smuggling network. These were similar to those being used by Iran and were transported to the Oak Ridge nuclear laboratory. “That information has percolated into the intelligence world and has given the US and the Israelis more information about how these machines operate,” says Hibbs. “With everything they know about the P1 centrifuge [the type used by Iran at present] the Americans would be very willing to put their knowledge to use.” He suggests that the Israelis, too, could have tested out Stuxnet. “The Israelis know everything there is to know about how to develop these kind of centrifuges. They have also got a lot of help from their friends in the US,” he says.
But how did Stuxnet get inside Natanz? Experts say the core industrial plant at the facility is not connected to the worldwide web, so the worm would need to be introduced by a hand-held USB device. The only non-Iranians who can enter Natanz are nuclear inspectors from the International Atomic Energy Agency, who regularly monitor the plant under Iran’s international treaty obligations. In December, Iran’s intelligence chief, Heidar Moslehi, accused the IAEA of sending western intelligence agents in the guise of inspectors to collect information about the nuclear programme. The IAEA denies Tehran’s accusations. “Our guys are insistent that they don’t know what caused Stuxnet or what impact it has had,” says an IAEA official. “All we do is monitor the nuclear material emerging from the plant. We don’t have access to machinery or look at the performance of equipment.”
Could the worm have been introduced by an Iranian scientist or technician? Hibbs is confident that the Iranian programme is penetrated by intelligence services. “Without deep penetration, that kind of sabotage is not possible,” he says. If such penetration has indeed taken place, then it probably went far beyond the ability to physically access Natanz. Whoever wrote Stuxnet also had access to confidential design information. The worm, say experts, would have been set off by Natanz’s specific attributes – including the speed of centrifuges and the type and number of technical components known as frequency converters. “Somebody had to have a spy in Natanz to have that specific information,” Albright says. “They had to know something about Natanz that the IAEA does not.”
Experts who have studied Stuxnet argue that its introduction into Natanz was an organised, government-backed effort that took months, if not years, of planning. But there is a divergence of views over whether Stuxnet has been a success. True, the US and its allies have expressed ever greater confidence that the day of reckoning over Iran’s nuclear programme is still some way off. It is not just Clinton’s talk of delay. Israel has long predicted Iran could have the materials for a nuclear bomb within a year to 18 months. But in December Moshe Yaalon, strategic affairs minister, said he believed Tehran was at least three years away. Then in January, Dagan, the retiring head of Mossad, predicted Iran would not build a nuclear weapon before 2015.
Yoel Guzansky, former head of the Iran desk on Israel’s National Security Council, argues that, taken as a whole, the covert campaign is now the “most effective tool for hindering Iran’s progress”. He praises covert operations for offering a smart compromise between sanctions and a military strike. “They can convince Iran to return to the negotiating table and moderate its position,” says Guzansky, who is now with Israel’s Institute for National Security Studies. “They buy precious time for diplomacy and sanctions and, if they really work, may encourage decision-makers in Israel not to rush to other options.”
But does this self-congratulatory mood reflect lasting disruptions in the Iranian programme or is it simply part of the psychological pressure being exerted on Tehran? While Iran’s programme may have been stricken, it has not been stopped. The country continues to produce 100kg of fuel grade material (low enriched uranium) a month and smaller amounts at slightly higher levels, although not yet weapons grade. Some experts argue that the impact of Stuxnet has been overestimated. “There is a strong debate inside Israel about where the clandestine efforts stand,” says Ronen Bergman, author of a forthcoming book about Mossad and a security analyst with Israel’s Yedioth Ahronoth newspaper. “There are people in Israel who claim that there is a peak moment for clandestine efforts aimed at stopping a scientific programme – and that that peak moment has already passed. The programme is advanced, and the Iranians are aware of the efforts [to undermine it].”
Bergman’s analysis is that Stuxnet was “a major success and a huge fiasco” at the same time. “The question is what is your purpose – to win the battle or win the war? The virus didn’t vaporise or disassemble, it was discovered, it infected non-Iranian machinery, and now the Iranians are aware that someone is trying. They are aware of the exact methods. It gives them an understanding of how the malware is able to jump between computers,” he says. According to Bergman, the story of Iraq’s nuclear programme offers a sobering lesson. In 1979, he says, a French warehouse housing key components for the planned Iraqi reactor at Osirak was blown up. The operation, presumed to be the work of Mossad, dealt a damaging setback to Iraq’s nuclear ambitions. Yet, not long after, Osirak was nearing completion, and in 1981, Israel took the intensely controversial decision of ordering an air strike.
There are other reasons to be cautious about the impact of cybersabotage. First, the US is worried that while the P1 centrifuges show signs of stress, Iran could deploy more efficient machines. North Korea, with which Tehran has traded on the black market in the past, has installed the more sophisticated centrifuges. Second, officials in Washington concede that if Iran is bent on developing a military application to its nuclear programme, it would probably be operating well away from Natanz and the prying eyes of UN nuclear inspectors. These suspicions were reinforced when western intelligence agencies revealed in September 2009 that Iran had been building a secret enrichment facility near the holy city of Qom. The facility is built in tunnels and experts believe it could provide an alternative for Iran if Natanz was bombed; or it could be used to enrich uranium to a higher, weapons-grade level. “Seeking a covert capability is by far the safest way for them to try to develop a nuclear weapons capacity,” says Gary Samore, Barack Obama’s top adviser on Iran’s nuclear programme. “I think we should assume that in the future Iran is going to try to proceed to build secret facilities.”
One senior western diplomat says that it is too early to make a judgment. “You don’t know whether you are winning until you have won. It is like throwing stones into a pond: you see nothing as the stones pile up under water, and then suddenly the first one breaks the surface.” For some experts, moreover, the sabotage campaign comes at a cost. Stuxnet may have concentrated its efforts on Natanz but it has also proved expensive for many industrial companies. And should Iran choose to respond, it could prove just the beginning of a dangerous cyberwar, to which neither the US nor Israel are immune. US civilian and military officials have noted that America’s key infrastructure – such as power and water plants – is frighteningly vulnerable to cyberattack. “It’s a dangerous technology,” says Frantz, the expert on the CIA’s sabotage efforts, referring to Stuxnet. “Releasing it does two things: it spreads the danger but it also opens the door for retaliation. And I think that cyberwarfare is the next big front.”
The new type of covert war that has ensnared scientists, unleashed dangerous viruses and sought novel ways of exporting faulty equipment takes the nuclear stand-off into uncharted territory. If effective, it buys the US and its partners time, postponing the day when they might have to decide between a conventional strike against Iranian nuclear facilities, with all the risks that it engenders for the Middle East, or acceptance of Iran becoming a nuclear-capable state. But every war has a cost, and in this mysterious world of intrigue and sabotage, no one knows yet what the real price will be. “There’s too much happening behind the scenes with black programmes, assassinations, sabotage of equipment, cyberattacks,” says Albright of Isis. “There’s a loss of accountability … There’s a sense that a green light has been given on hitting the Iranian nuclear programme. But who is going to take the lead on establishing ground rules on what is too much? Where does it end?”
To comment on this article, please e-mail firstname.lastname@example.org