Mannheim is a medium-size city in southwestern Germany best known for the large American military base on its outskirts. The city’s courthouse was built in the 1960s, and the years have not treated it well. The exterior is poured concrete, with the second story covered by dull, rusty steel. Inside, the courtrooms, with their popcorn-stucco walls and yellowed ceilings, can be described most charitably as utilitarian. All in all, it was an inauspicious location for an event of historic importance. But at ten o’clock on the morning of March 17, 2006, the trial of a suspected member of A. Q. Khan’s nuclear mafia started in courtroom two, which was the size of a high school gymnasium.
Sitting at the defense table that morning, arms folded across his chest and a binder open in front of him, was Gotthard Lerch. At sixty-three and about six feet tall, he looked fit and comfortable, particularly for someone facing the possibility of spending the rest of his life in prison. Perhaps Lerch appeared at ease because he was confident that the case against him had been fatally weakened. Infighting among the governments of several countries was threatening to deprive the prosecution of key witnesses against Lerch. Or perhaps Lerch wasn’t showing any sign of worry because he had been through it all before: The German government had tried twice to prosecute him, once in the late 1980s and again in 1992. Each time, he had walked away a free man and resumed selling nuclear technology.
This third attempt was in trouble before the judge called the proceedings to order that Friday morning. For months the German prosecutors had been struggling to gather evidence against Lerch. The intelligence agencies that presumably knew the most about his role, the CIA and MI6, were providing no information. The prosecutors had even run into a stone wall with their own intelligence agency, which had been cooperating with its counterparts at the CIA and MI6 but was not inclined to share any information with colleagues in its own government. Requests for help from governments of Malaysia, South Africa, and the United States had fallen largely on deaf ears, too. Prosecutors had traveled to Malaysia and interviewed B. S. A. Tahir, who had repeated his accusations that Lerch had been involved intimately in procuring and shipping centrifuge technology to Libya for its nuclear weapons project. But the Malaysian government was refusing to send Tahir to Mannheim to testify. Instead they provided an affidavit confirming his statements, which was inadmissible because German law entitles a defendant to cross-examine his accusers. Similar problems arose when prosecutors tried to arrange for Gerhard Wisser and Daniel Geiges to come from South Africa to testify. Still smarting over the unilateral interview conducted by the Germans during the Pelindaba review, the South Africans were refusing to make either man available. They did offer, however, to have Wisser and Geiges respond to questions through a video linkup from a South African courtroom. The prosecutors were still negotiating with both governments, but prospects looked dim.
The prosecutors also faced problems related to the deal in which the Swiss had agreed to send Lerch back to Germany for trial in exchange for Urs Tinner. The Swiss had required the Germans to drop the treason charge because it was regarded as a political crime and the Swiss constitution prohibits extradition for political charges. Dropping the most serious charge against Lerch meant the case was transferred from the federal system to the local court in Mannheim, where prosecutors had less experience with complex international crimes. The Germans had also failed to obtain a guarantee from Switzerland that Tinner would be returned to testify against Lerch.
Tinner was expected to be the star witness. During his months in a German prison, he had been coddled by his jailers. The first time that Roman Boegli, his lawyer, had gone to see Tinner in his German prison, they met in a restaurant reserved for police officers. His client was wearing jeans and a T-shirt. He seemed on friendly terms with the police. “He was a king in Germany,” Boegli said. “They liked him there. I believe he was used by the German secret service. They never tried to accuse him of anything, but they got as much information from him as they could.” The truth was, the Germans had been treating Tinner well because they needed him for their own prosecution of Lerch.
But now the star witness was looking like a no-show. Just as Tahir’s affidavit was of limited value without his presence in the courtroom, Tinner’s statements also would be almost useless if he were not on the stand to testify. For reasons that the Germans could not understand, the Swiss were refusing to send their prisoner to Mannheim.
The prosecution needed to lay out an understandable case that demonstrated that Lerch had knowingly helped the Libyan nuclear program through specific actions. The defense strategy, on the other hand, planned to exploit the mystery surrounding the case. They would argue that their client could not get a fair trial when the witnesses against him were hidden away in foreign countries. Further, the defense lawyers were contending that the intelligence agencies of the United States and Germany were conspiring to conceal the extent of their involvement in the Khan network and how they had manipulated evidence against their client. Without information from the CIA and other intelligence services, the lawyers maintained, the court could not understand how their client had been sucked into an operation concocted by intelligence agencies. Lerch, they contended, was a victim, not a perpetrator.
On March 17, most of the opening session was taken up by arguments from Lerch’s lawyers that they had been denied full access to secret material used against their client. Even the small amounts of evidence disclosed to them was classified and they had been forbidden from copying or taking notes. The lawyers said they weren’t sure what they could discuss with Lerch or what they could bring up in court.
“I don’t know what I’m allowed to say,” Gottfried Reims, the lead defense attorney and a man given to dramatic statements, complained to the presiding judge. “Help me.”
The judge, Michael Seidling, said he had asked the German intelligence service how to handle the classified information, but that he had not yet received any instructions.
Reims’s cocounsel was sitting at the defense table with a laptop computer and a portable printer, furiously typing away. Within minutes he produced a motion demanding that three of the six judges hearing the evidence be removed because they had refused the defense access to the secret material, claiming that it was classified and too sensitive to be disclosed to anyone without the proper security clearance. Seidling had no choice but to adjourn the trial until the matter could be resolved.
When the trial resumed near the end of April, Seidling was still on the bench and the chief prosecutor, Peter Lintz, was finally ready to lay out the case against Lerch. Unfortunately for the prosecution, the young attorney was no closer to having any of his key witnesses show up in court to supply the vital corroboration of his accusations. With broad strokes, he described how A. Q. Khan had agreed to provide Libya with the capacity to produce an arsenal of nuclear weapons, from enriching uranium to assembling warheads. “To produce and supply the goods needed by Libya, Khan used a circle of proven helpers who had already been of service to him for his own centrifuge program,” Lintz said. Among the helpers he identified were Tahir, British businessman Peter Griffin, Urs Tinner, and Gotthard Lerch. “It was clear from the start to all those involved that the goods to be produced were for a gas centrifuge project for the purpose of highly enriching uranium in Libya,” Lintz told the court.
The prosecutor argued that Lerch had played a central role in the Libyan project, organizing the training of Libyan technicians and arranging for a massive feed-and-withdrawal system to be constructed in South Africa for shipment to Libya. In exchange for his work, Lintz said, Lerch had been paid $34 million, about half of which was profit. His actions, said the prosecutor, “endangered peace between peoples, did considerable damage to Germany’s foreign relations, and risked seriously threatening Germany’s external security.”
A conviction relied heavily on proving that Lerch attended specific meetings at which the Libyan project was discussed and took specific actions to supply designs and equipment. Lintz had paperwork that had been gathered by investigators, and he had the statements of people like Tahir, Tinner, and Meyer. Live witnesses were needed to explain how the complicated scheme had worked and to link Lerch directly to the criminal activities. Even the delay until the end of April had not solved that problem for Lintz. So far, only one witness associated with the shipments had agreed to testify, and he would cause more damage than good.
In mid-May, Peter Griffin took the train to Mannheim from his home in Bordeaux, France. A British businessman who had retired to the south of France, Griffin had spent years denying accusations of involvement in illegal exports of nuclear technology to Pakistan and he had never been charged with a crime. Still, Griffin remained cautious. When the German prosecutor asked him to testify in Mannheim, he insisted that the court provide him with a letter promising that he would not be taken into custody by the Germans. “During the said period of time it is not permitted to apprehend or arrest the witness for a possible involvement in the charges specified in the above-mentioned indictment,” said the letter.
Griffin arrived at court shortly after two o’clock on a Tuesday. The judge asked him to wait in the lobby until it was time for him to testify. As he sat on a bench with Steve Coll, the writer for the New Yorker who was covering the trial, the two men watched as a stream of spectators from two other courtrooms gathered in the lobby and chatted about their cases. One of those trials involved a prostitution ring and the other was a double murder arising from a love triangle. “This is nothing, then,” Griffin said to Coll. “This is only the Apocalypse.”
The businessman spent two hours on the stand, focused on proving his own innocence and saying little about Lerch. Griffin testified that he had been duped by Tahir, who had never told him that any of the shipments through his company were destined for Libya or any other nuclear program. At one point, he said that Tahir had forged invoices from his company, Gulf Technical Industries. Worse for the prosecution, Griffin said on the stand that Tahir’s statements should be regarded as unreliable; the testimony brought a smile to Lerch’s face. The case seemed to be slipping from the prosecution’s grasp. It looked as if Lerch would walk free again.
Lerch’s lawyers certainly seemed unconcerned. During breaks they often gathered in a corner of the hall outside the courtroom where they could smoke and banter. One afternoon not long after Griffin’s testimony, Reims was talking loudly about articles he had read that described how George Tenet had boasted that the CIA’s penetration of Khan’s network was so thorough that agents had even searched the Pakistani’s house.
“Let us presume that is true,” Reims said to no one in particular, but loud enough for anyone to hear. “Then let Mr. Tenet tell us what he found out in his brave operations, in Khan’s living room and at his company. What is the role of the CIA? What is the role of the other secret services? Are all these people we hear about here, whose payroll are they on?”
On July 27, Judge Seidling delivered the verdict. He announced that the judges had determined that Lerch should be freed. The lack of foreign cooperation and absence of key witnesses meant that the defendant could not receive a fair trial, said the judge. However, the court stopped short of exonerating Lerch. Instead they left open the possibility that the trial might resume after a fuller disclosure of evidence. Lerch’s bail was set at $5 million. Given the difficulties surrounding his extradition from Switzerland, the court ordered Lerch to remain in Germany until a decision was made on whether to reopen the case against him. For a third time, Lerch walked away.
A SIMILAR DRAMA WAS PLAYING out in South Africa. Prosecutors there who were building the case against Wisser and Geiges were confronting difficulties getting cooperation from the American government. The South Africans had the hard evidence—eleven shipping containers of equipment that was intended for a uranium enrichment plant. They had the cooperation of a key witness, Johan Meyer, who was prepared to testify that he had built the feed-and-withdrawal system on instructions from Wisser and that they both knew it was destined for a Libyan nuclear plant. But the prosecution wanted to strengthen the case with testimony from American experts. They were the ones who could provide the context for the Libyan program and describe the scope and danger of the Khan network. But the Americans were threatening to withhold all help unless the South Africans agreed to conduct the entire Wisser and Geiges trial behind closed doors, which contradicted South African law.
The American attitude was confounding and contradictory, pulled in two distinct directions. “As a general rule, we wanted to see these individuals prosecuted,” Robert Joseph, the Bush administration’s undersecretary of state for arms control who was involved in negotiations with many countries over prosecutions of network participants, said later. “I have this old-fashioned notion that people ought to be punished for crimes, and we need to show the world that if you engage in this illicit but lucrative business, there will be a punishment. It would have to be something that demonstrates a deterrent effect. If there is no downside, why wouldn’t you do this?”
But Joseph’s old-fashioned notions of justice ran into opposition from the CIA and its advocates at the White House and elsewhere in the administration. Certainly there were genuine fears among the American officials that public trials could jeopardize continuing intelligence operations. For many people those risks outweighed the consequences of failing to prosecute network participants. They were more determined to protect the illusion that shutting down the Khan operation was a huge intelligence victory by keeping the Tinners out of a Swiss court, even if it meant allowing some nuclear traffickers to escape punishment.
By the end of 2006, three years after the seizure of the BBC China, only one person had been convicted in connection with the nuclear network. He was a minor figure, Henk Slebos, who had no direct role in supplying Iran or Libya.
Khan’s nuclear mafia was widely acknowledged as the most dangerous private proliferation ring in history, helping Pakistan build its atomic arsenal and providing critical assistance to at least three other countries—Iran, Libya, and North Korea. But in trying to deal with crimes that threaten international security, prosecuting the criminals who sell nuclear technology and know-how requires a balancing act. On the one hand there’s the goal of putting individuals in jail, and on the other hand is the goal of protecting the sources and methods at the heart of intelligence gathering. Too often, however, the CIA and other intelligence agencies claim to be protecting sources and methods when they are really protecting bad judgments and operational errors.
In Khan’s case, the balance tipped to the intelligence side of the ledger from the start. As far back as 1975, he could have been stopped when the Dutch secret service first discovered that he was stealing centrifuge designs from Urenco. But the CIA intervened, pressuring the Dutch government to allow Khan to remain free so that his activities could be monitored. Initially the intelligence imperative allowed Khan to operate. The Americans and other intelligence agencies watched over the years as he played a central role in building his country’s nuclear arsenal. But Khan’s way was also smoothed by a succession of American leaders. They traded strict standards against nuclear proliferation for other goals, starting with the Carter administration’s determination to ignore Pakistan’s nuclear ambitions in order to maintain the country’s assistance against the Soviets in Afghanistan in late 1979.
The same sort of permissiveness resurfaced in the Bush administration after Khan’s confession in 2004. The United States never complained when the Pakistani government treated Khan with what Leonard Spector, a nonproliferation expert in Washington, referred to as “unseemly leniency.” Instead of insisting on questioning Khan to get to the bottom of who received what technology from his nuclear network, the administration allowed Pakistani authorities to isolate Khan from any outside interrogation as the price of Islamabad’s support in the fight against the insurgency in Afghanistan. And instead of providing the Americans or the International Atomic Energy Agency with a road map to undo the damage he caused, Khan was allowed to boast on a Karachi television show about the extensive assistance he provided to Iran’s nuclear program and why he did it. “If Iran succeeds in acquiring nuclear technology, we will be a strong bloc in the region to counter international pressure,” Khan said. “Iran’s nuclear capability will neutralize Israel’s power.”
The same shortsighted strategy was behind the American government’s refusal to cooperate fully in prosecuting the individual participants in the Khan network. Washington should have been providing all of the expert testimony and evidence available to go after every individual and company associated with a nuclear black market that had provided the world’s most lethal technology to countries like Iran, North Korea, and Libya. Instead, the intelligence imperative was driving U.S. policy. Protecting the chess pieces in the game of espionage outranked punishing Khan and his associates and sending a powerful message of deterrence to future proliferators. The intelligence community, where information too often becomes the end goal, failed to act at critical junctures throughout the history of the Khan network. It failed to stop Khan in 1975, and it compounded that failure by not detecting the extent of his success in building Pakistan’s atomic bomb. It also failed to detect the extent of the network’s proliferation activities, both in terms of geography and technology. And, in what may turn out to be the biggest blunder in counterproliferation history, the CIA and other intelligence agencies missed Iran’s acquisition of the hardware and know-how to make an atomic bomb. Even the best-intentioned policymakers had not been able to overcome a flawed intelligence strategy that spanned three decades.
The failure to convict Lerch in Germany and the difficulties confronting the prosecution of Wisser and Geiges in South Africa were stark evidence of the ongoing cost of the American position. But the most extreme example of this wrongheaded strategy was the widening campaign of the CIA and its backers to prevent the exposure of the Tinners in Switzerland. Senior administration officials were arguing that the risks to continuing intelligence operations were too high. No one was asking about the long-term costs of burying the truth about the damage done to world security by the Khan network generally and the Tinners specifically. Such costs extended beyond the growing threat posed by a nuclear Iran, because no country develops the atomic bomb in isolation. The ripples of Iran’s success—a success that would have been impossible without the help of Khan and his associates—will be felt across the Middle East as its neighbors are forced to confront a new nuclear reality.