by Shane Harris
Libertarians and privacy activists reacted to these ideas with a mix of outrage and apoplexy. The government was coming at them from two sides—setting standards for the architecture of the network and becoming the traffic cop for everything that moved on it. An atmosphere of hostile skepticism permeated all conversations about the new telecom law and the Clipper chip.
Beginning in August 1994, senior law enforcement officials sat down for meetings in Washington with a coalition of more than four dozen activist groups and technology companies, including the biggest telecom provider of all, AT&T, as well as IBM, Microsoft, and the U.S. Telephone Association, which represented more than 1,200 local phone companies and the so-called Baby Bells, which were created after the government breakup of AT&T in the early eighties. None of these companies wanted to see draconian measures suffocate the spirit of invention. But they also couldn’t object to the government’s legal and security imperatives. Officials had to be able to monitor communications legally in order to protect the country against all kinds of threats, known and unknown. Somehow the two sides had to compromise.
The meetings featured intense, nitty-gritty debates over the technical aspects of the proposed law. The government wanted guarantees that the telecom system would never mature beyond the reach of its wiretaps. Some companies saw this as heavy-handed regulation, and a number of telecom officials shared the activists’ belief that the government was after a permanent backdoor into the phone system. The negotiations helped to dampen the suspicions somewhat, however, and the talks went forward, because no one in the room disagreed with the fundamental premise that the government had the right to wiretap.
The activists had a bottom line. They only would agree to solutions for known problems. The new law mustn’t be written to address future gaps. If the FBI had difficulty accessing the public telephone network, then the law would address only that public telephone network. The activists wanted to set a philosophical guideline, but they also had another strategic goal, one they felt must be defended at all costs: They wanted to keep the government’s hands off the Internet.
In 1994, the net was so new that its future was still notional. Only a handful of subscribers were surfing online through providers such as America Online and Prodigy. The first Web browser had just been released, in a beta test version. But the activists knew that the Internet was a free space, an environment fueled entirely by innovation, and that it held as yet unimaginable promise. The government never would agree willingly to stay out of that space. If that’s where the world was moving, that’s where the cops and spies wanted to be. Needed to be.
During one of the meetings, David Johnson, a lawyer who had helped to craft a landmark electronic privacy act in 1986, demonstrated why the Internet was destined to be such a valuable intelligence resource. He held up a glass jar full of rocks and asked the room, “How many of you would say this jar is full?” Most everyone agreed that it was. Then Johnson took a fistful of pebbles and dropped them into the jar. They tinkled down through the rocks, finding resting places in the empty spaces. Then he poured in some sand. As it cascaded into the empty spaces, he told the onlookers that the sand was like the unseen, seemingly insignificant “transactional data” that traveled on the network. It held routing information for a text-based message like an e-mail—where the message came from, where it was going, what path it followed through the light-filled pipes. The transactional data also contained a series of digits that composed an Internet address, a unique location in cyberspace. It was starting to sound like RR326. A beacon in the electronic storm. This information someday would be of enormous value to the government, Johnson said, just as phone call records already were. The transactional data were small but meaningful, just like the tiny grains of sand that filled up the jar.
Two influential Democrats had attended the meetings, and they came out convinced that the government had to be restrained. Senator Patrick Leahy, from Vermont, and Congressman Don Edwards, whose district included Silicon Valley, formally introduced new telecom legislation and declared that it would apply solely to the public telephone network. The law specifically exempted “information services,” which all the parties agreed included Internet companies and electronic-messaging technologies.
The Communications Assistance for Law Enforcement Act passed in the closing days of the 103rd Congress, two weeks before Republicans won control of both chambers in November 1994. CALEA (pronounced kuhLEEuh) would let the industry set its own standards to meet the Justice Department’s needs. The department could list its surveillance requirements, but the act let companies decide how to build their own equipment. Officials could petition the Federal Communications Commission if they felt that the companies weren’t fulfilling their obligations under the law.
The government could have stopped there. It could have settled for secure access to the phone networks and kept electronic surveillance intact. But the government had never intended to cede cyberspace to technicians and activists, any more than to coke dealers and hitmen. The law was merely a new starting point. Just weeks after Clinton signed CALEA into law, in January 1995, the raft of compromise sprung its first leak.
Law enforcement officials had made a calculated decision to leave the negotiations and then quickly regroup when it came time to set the standards for implementing CALEA. That’s when the government would be able to present its list of technical requirements for the phone system. They’d lay out in precise detail what kinds of surveillance powers they needed, and what the telecoms were expected to provide. The privacy advocates were dragging their feet in the negotiations, FBI officials concluded. Delay would invite more debate, probably hearings, and possibly a less favorable outcome. Freeh made the political call: Let’s take what we can get here.
In early 1995, the Justice Department issued its list of requirements for wiretapping, known as the “punch list.” Telecom executives and their lawyers were dumbfounded. They thought the requirements exceeded reasonable needs for the government and violated the spirit of negotiation from the previous summer. Al Gidari, a lawyer who represented the wireless industry, was among the first to see the FBI’s punch list during a standards-setting meeting in Vancouver, British Columbia. He thought it looked like the Cadillac of wiretaps.
Everything that the bureau could think of was on that list. It would be exponentially more difficult for the companies to comply with these demands, Gidari thought. This went beyond simple wiretapping. There were highly technical processes involved in the kinds of sophisticated surveillance that the government now imagined doing. Their list eventually grew to include real-time geographical tracking of mobile phones; the ability to monitor all parties in a conference call regardless of whether they were on hold or participating; and “dialed digit extraction,” a record of any numbers that a subject under surveillance punched in during a call, such as a credit card or bank account number.
The FBI wouldn’t be the only user of this data, of course. If the system were engineered to provide this fine level of data legally, the intelligence agencies would want it too. Again, the premise of wiretapping wasn’t in question. But now serious doubts were rising about whether the government’s consumption of information would stop with a simple phone call.
The standards meetings were tense and awkward, and the sides were unevenly matched. On one side of a conference table sat a dozen FBI agents, humorless, wearing neat blue suits—buttoned down and ready to roll over their opponents. On the other side, a disheveled, laid-back, quirky bunch of network engineers representing the telecom carriers and the equipment makers. It fell to them to tell the FBI, No, you can’t have this. These are precisely the kinds of terms that the law says you cannot dictate.
The meetings bore no resemblance to negotiation. Some of what the government wanted looked illegally broad. Much of it just seemed excessive. The level of government surveillance was so low at that time that some questioned why the FBI wanted multifaceted access. In 1994, federal and state authorit
ies were running only 1,154 wiretaps nationwide, mostly for drug investigations, at an average cost of fifty thousand dollars each. Intelligence wiretaps under FISA were rarer. As Gidari saw it, the government was asking carriers to design a nuclear rocket ship for a flight from Chicago to D.C. It seemed to him that officials thought there was no limit to the expense the companies should bear in order to save a life.
When the companies protested that the surveillance specifications were too complex, and that they’d add extra cost and force the equipment makers to adopt rigid technical standards, the government questioned the companies’ loyalties. In one meeting Gidari watched as FBI and Justice officials slammed their hands on a table and screamed, “You’re unpatriotic! What do you want to do, help the criminals?”
Unable to get everything it wanted from the punch list, the government petitioned the Federal Communications Commission, as the law allowed. Industry lawyers joined up with activists and told the commission that the negotiations had deadlocked because of the government’s intransigence. Officials had made “unreasonable demands . . . for more surveillance features than either CALEA or the [analog] wiretap laws allow,” they argued in legal papers. But the panel members were inclined to side with the government. The tussle dragged on for two more years, ending up in a court of appeals, where the FCC finally was overruled. The commission took up the matter again, and this time struck more of a balance between what officials wanted and the companies were prepared to give.
It was a temporary peace. Intelligence and law enforcement officials continued to eye changes to the law and made plans to go after the biggest prize of all—the Internet. Half a decade later the moment would arrive when a national crisis proved that the system had gotten out of their hands. Then, there would be no stopping them.
The FBI and the NSA had managed to survive the digital wave. The Clipper chip was ultimately junked, after privacy activists and a number of apolitical technology experts questioned its usefulness. But by the midnineties, the NSA, along with the rest of the government, could be sure that the networks would never be built beyond their eavesdropping capacity. That set a baseline for the information age. And it was the catalyst for a new, constant tension between the government and the governed that would last well into the next decade. The balance of power in a new world had been reset. And into those strange, uncharted waters, an aging sailor pointed his bow.
CHAPTER 6
THE GENOA PROJECT
Now nearly sixty, John Poindexter had assumed the curious existence of a Washington pariah: not really gone but officially not still around.
It was 1995, a decade since the Iran-Contra affair. And it wasn’t that people were afraid to be seen with him. In a closet the size of Washington, skeletons never stayed put for long. Poindexter’s friends and associates even could make light of the “scandal,” a word he hated hearing people use, that sullied Reagan’s twilight years and ensured Poindexter’s status as an ignominious footnote. At a luncheon hosted by the Naval Academy, some time after the controversy had dissipated, Poindexter was seated at the same table as Tom Clancy, the bestselling master of the cold war sea story. His most recent book, Clear and Present Danger, centered on a renegade national security adviser and his deputy who wage a secret war in Central America and, in a climactic scene, try to delete the e-mails detailing their exploits. “Admiral,” Clancy said across the table, “I just want you to know that I did not base the character in my book on you!” Poindexter smiled and nodded.
He had appealed his five felony convictions. His lawyers, paid for by friends and political supporters, argued that the testimony Poindexter had given to Congress under a grant of immunity was improperly used against him at trial. In November 1991, a panel of three federal judges agreed and overturned the jury’s verdict. It was a quiet vindication. He had occasion to thank George H. W. Bush personally for not pardoning him: “I’m glad I did it on my own,” he told him a few months later, at a White House ceremony honoring former president Reagan with the Medal of Freedom. Reagan, now in his early eighties, had testified for Poindexter in his criminal trial. But when he reached a now frailer man in the receiving line, he could see the disease was in bloom. “Mr. President, I’m John Poindexter,” he said, reintroducing himself.
Plenty of people remembered Poindexter fondly, enough to erase any regret that his prodigious confidence and sense of righteousness couldn’t. About a year after his conviction was overturned, Poindexter was stopped in his driveway by a man walking with his two daughters. “Aren’t you Admiral Poindexter?” he asked. “Yes,” he replied.
“Well, I just wanted to thank you and to tell you what a great man I think you are.”
One of the daughters was confused. “Daddy, who is this?” she inquired.
“Well, this is Admiral Poindexter.” It didn’t register. “What does he do?” she asked.
The man thought a moment. “He helped the country. He used to work for the president.”
People often remarked to Poindexter that the tumult of the past years didn’t seem to have taken a toll on him. This was why. Moments like this sustained him.
Poindexter’s contempt for Congress was not diminished by the spectacle of televised testimony before the joint committee investigating Iran-Contra. The lawmakers who had hauled him before the cameras hoping he’d hand them a smoking gun were incensed to watch him fall on his sword. Reagan had known nothing of the diversion of funds to the Contras, Poindexter said. He had authorized the operation himself and deliberately kept the president in the dark, a gift of plausible deniability.
“The buck stops here with me,” he intoned.
Years later the image of the admiral sitting placidly before the lawmakers, seemingly disinterested as he stared through his trifocals, still burned on their minds with the slow intensity of the tobacco embers that coughed smoke into the hearing room from Poindexter’s pipe. He smoked throughout the proceedings, lighting the bowl with a silver Zippo that was mailed to him by the owner of the company, an ardent political supporter.
The buck really had stopped with him. An independent investigator, a wave of journalists, and the president’s attorneys spent years scrubbing every inch of documentary evidence on Iran-Contra. They conducted hundreds of hours of interviews and found no evidence that Poindexter had lied abut the president’s role. That hardly exonerated either man; indeed, for many it reaffirmed just how unforgivably out of touch Reagan had been all along, and that Poindexter really was the rogue people had long described. But Poindexter’s fall had saved Reagan’s presidency. It was, in his eyes, a final act of service. Had Congress determined that Reagan knew of the diversion, and authorized it, they would have pursued impeachment.
The turmoil of that bitter period failed to diminish Poindexter’s politics, his pride, or his passion. But it did take away his status. He had staked his career on trying to change a system from the inside, and he had failed.
But then a window opened. In late 1995, one of Poindexter’s old academy classmates surfaced with a tempting invitation. He was working for a technology firm in Northern Virginia called Syntek. The business was started by another legendary Navy man, Reuven Leopold, who for most of the 1970s was the Navy’s technical director for ship design. Leopold had spent much of his career fighting wars from the drawing board; he led the construction of twenty-one classes of vessels, including the famed Spruance destroyers. But now, with the Navy cutting back its shipbuilding budgets, Leopold wanted in on the next big thing: information technology. And he was eyeing a promising new program in the Pentagon’s research and development office about which he thought Poindexter might have some special insight.
It was called Spinnaker, and it caught Poindexter’s attention for more reasons than its nautical nomenclature. For starters, he was a great fan of the agency running the program—the Defense Advanced Research Projects Agency. Housed in an unassuming office building across the Potomac River in Northern Virginia, DARPA was the Pentagon’s futurist brain
trust. Its researchers had built the first “Internet.” They developed stealth technology that turned aircraft invisible. DARPA tackled what researchers called the “hard problems,” challenges for which a solution was years, perhaps decades, away. It was Poindexter’s kind of place, reminiscent in some ways of the Whiz Kids’ shop under McNamara but on a grander, far more ambitious and systematic scale. DARPA’s province was one of risky ventures with revolutionary potential. The agency embraced failure as a necessary part of an idea’s evolution.
Spinnaker fell under the aegis of DARPA’s Information Systems Office, a unit whose mandate was as broad as its title. Information technology, particularly computer software, was altering the course of human society in ways that most people couldn’t quite articulate but still understood were fundamentally important. But the fervor hadn’t seized the ranks of the U.S. government, which remained locked in a technological dark age. The wave emanating from Silicon Valley hadn’t yet swept over Washington.
A lot had changed in the decade since Poindexter left government. Now the technology existed to start building more sophisticated versions of the intelligence fusion centers and sharing systems that he had envisioned at the White House. Most of it was manufactured by companies, not by federal agencies. New products and concepts portended an information revolution, and DARPA wanted to harness it.