The Wizards of Langley (49 page)

At the analyst level, there were also concerns. Patrick Eddington joined the CIA in 1987 after graduating from Southwest Missouri State. After attending the agency’s National Imagery Analysis Course, he reported to NPIC in February 1988. Over the next seven years, he served in both the Priority Exploitation Group at Ft. Belvoir and the Imagery Exploitation Group at the Washington Navy Yard, as well as with the intelligence directorate’s Office of Scientific and Weapons Research. Over those years, his interpretation efforts had been focused on imagery of the Soviet Union and Iraq.
⁵⁵

He later wrote that NIMA was “derisively referred to as ‘the Enema’ by NPIC analysts” and “was seen as little more than a power grab by the Pentagon, orchestrated by Deutch to gain a monopoly over the national imagery system.” “No self-respecting imagery analyst,” Eddington wrote, “wanted to become a ‘human photomat’ producing a mountain of meaningless briefing boards for the Joint Chiefs of Staff or any other Pentagon ‘customer.’”
⁵⁶

Objections also came from the vice-chairman of the Senate Select Committee on Intelligence, Robert Kerrey (D.–Nebr.), and the House Permanent Select Committee on Intelligence. The primary concern was that as a result of the transfer of NPIC personnel from the CIA to DOD, imagery support to national policymakers would suffer in order to support the requirements of military commanders. Although the opposition was unable to block the creation of the new agency, the Senate Select Committee on Intelligence did persuade the Senate Armed Services Committee to amend the legislation creating NIMA. Thus, the final legislation stipulated that the DCI retained tasking authority over national imagery systems, and that the Secretary of Defense must obtain the DCI’s concurrence before appointing the NIMA director or note the DCI’s lack of concurrence before recommending a candidate to the president. In addition, the Armed Services committee agreed to modification of the National Security Act to explicitly state NIMA’s responsibility to provide intelligence for national policymakers.
⁵⁷

When NIMA came into being on October 1, 1996, it incorporated all the elements mentioned in the late November statement as well as the Air Force’s Defense Dissemination Program Office, which disseminated
satellite imagery, and the CIA’s Office of Imagery Analysis—effectively removing from the CIA any responsibility for imagery interpretation. Heading the agency was Rear Admiral Joseph Dantone, with former NPIC director Leo Hazelwood as deputy director.
⁵⁸

At approximately 10 a.m. on December 20, 1996, a Titan 4 rocket, with about 3 million pounds of thrust, blasted off from its launchpad at Van-denberg AFB in California. To some extent, there was no mystery as to what was sitting atop the rocket as it headed for outer space. NRO spokeswoman Katherine Schneider acknowledged that the payload was a reconnaissance satellite, the first time NRO acknowledged such activity within two decades of its taking place. Other details, such as orbit and mission, remained classified.
⁵⁹

However, it was soon clear to observers that the payload was the third and final piece of the advanced KH-11 constellation, satellites whose numerical designations were 2104, 2105, and 2106. It was operating in an orbit of 155-by-620 miles and at an inclination of 98 degrees. Its two predecessors had been launched on November 28, 1992, and December 5, 1995, also using Titan 4 boosters, and were operating in similar orbits.
⁶⁰

Although in late 1997 they were “owned” by the NRO, they were the legacy of the Office of Development and Engineering and the work of Robert Kohler, Julian Caballero, Ed Nowinski, and Jeffrey Harris. Each satellite, often referred to as an Improved CRYSTAL, was enhanced in four ways: It had greater resolution than the KH-11 models (better than six inches); it was better able to perform area surveillance missions; it carried an infrared imagery capability, code-named DRAGON, which allowed it to image targets at night; and it carried the Improved CRYSTAL Metric System that allowed “fiduciary marks” to be incorporated in imagery—marks that enhanced the ability to use the imagery for precision mapping.
⁶¹

By the time the third advanced KH-11 was launched, its two predecessors had provided the intelligence community and its customers with high-resolution imagery of an array of global targets—a massacre site in Bosnia, Chinese military deployments, Indian preparations for a nuclear test in 1995, Iraqi military facilities, a Libyan chemical weapons plant, and North Korean missile test preparations.
⁶²

In the mid-1990s, the vision statement of the Office of Research and Development promised that “when the future arrives ORD will have been there.” Few people expected in 1995 that the office had a very limited future. But in addition to creating three new offices in 1996, in 1998 Ruth David abolished ORD—the only surviving original office.

Part of ORD’s mission had been to “fail”—to separate ideas that would work from those that wouldn’t before millions of dollars were wasted. But it had a number of outright successes and made contributions to the technical collection, clandestine collection, and analytical activities of the CIA and intelligence community—as well as to other parts of the government and private sector.

In 1984 and 1985, ORD sponsored the development of a “problem structuring aid” that was to be a tool to help researchers organize their ideas. The project evolved into a research program producing dozens of technical publications, and it produced one of the first hypertext systems. Although this hypertext system never became a product, its features were incorporated into software available to millions of owners of personal computers.
⁶³

In approximately 1985, the CIA decided to invest in the development of new tools and techniques to support the analysis of data. ORD identified information retrieval as a research topic of importance to the intelligence community that was receiving relatively little government funding. The office sponsored a project that, according to the agency, was so successful in achieving more accurate information retrieval that it was copied and incorporated into several commercial systems for automatic screening and sorting of news wires.
⁶⁴

Beginning in 1987, ORD led in the development of image perspective transformation modeling and visualization tools. The system took overhead imagery and, through use of image modeling and rendering tools, warped the images to appear as if the perspective were on the ground. This system proved extremely useful in the civil engineering and urban planning fields to create “what if ” scenarios. The operator could artificially insert and visualize new buildings or facilities on empty lots, or place new roads or interstates into a scene to see the impact on local communities. 65 It is also used to familiarize intelligence officers with an area place new roads or interstates into a scene to see the impact on local communities.
⁶⁵
It is also used to familiarize intelligence officers with an area where they will be operating, or arms control inspectors to become familiar with a facility to be inspected.

A joint project conducted by ORD and a component of NSA’s research and development directorate developed natural language processing (NLP)—which permitted computer processing of cables arriving at the CIA’s Counter-Terrorism Center to weed out irrelevant ones. Cables that took people one to three hours to sort through took a computer a mere ninety seconds. NLP could be used to locate documents containing the type of information the user desired or locate specified information from within a text. It could also be used to automate the construction of databases. By June 1995, it was being used by the Drug Enforcement Administration (DEA).
⁶⁶

Facial recognition, another ORD-developed technology, found use outside of the intelligence community by 1995. The facial-recognition program could be used to identify an unknown person against a set of known people. The technology was created to address a problem common to many agencies—databases with large collections of images, sometimes numbering in the millions, and collateral information. ORD wanted to be able to automate the process of identifying individuals using photographs and other existing databases. By mid-1995, users included the Immigration and Naturalization Service, the FBI, the Customs Service, and the DEA. In mid-1995, the INS arrested the first criminal to be identified by facial recognition, a convicted rapist who crossed from Mexico into Texas. The facial-recognition system enabled the INS to obtain positive identification of the individual within thirty minutes.
⁶⁷

More important, according to Gene Poteat, was ORD’s role in achieving major breakthroughs vital to carrying out the overhead reconnaissance mission. The breakthroughs, which he would not specify in detail, included advances in optics, imagery interpretation, and a unique fusion of collection and analysis. ORD’s contributions in that area had often been the work of the office’s Applied Physics Division, which was funded through the Reconnaissance Technology budget of the NRO. ORD’s accomplishments led Julian Caballero, sometime after he became OD&E head in 1985, to permit ORD to compete with OD&E for funds for proposed reconnaissance technology projects on an equal footing.
⁶⁸

David explained that her decision to abolish ORD and establish a small Technology Investment Office (subsequently known as the Investment Program Office, IPO) in its place was based on her belief that because the directorate subcontracted a vast amount of its research and development work (as it had from its inception), with little being done in-house, it did not make sense to maintain a centralized R&D organization. As a result,
she decided to distribute ORD’s personnel among the offices it supported, while protecting the long-term budget by creating the IPO. The investment office was to ensure that research and development funds were spent on long-term research projects and not shifted by individual offices to cover areas of immediate concern. It was, she acknowledged, a “controversial decision.” There were “people who hated it,” as well as people thought it a great idea, according to David. Not surprisingly, ORD members hated it, believing it devalued their contribution.
⁶⁹

John McMahon noted that the abolition was a “sign of the times” and a reflection that commercial technology was often far ahead of the gov-ernment’s—a dramatic change from 1962. Yet he believed ORD’s abolition would “narrow the agency’s knowledge into a lot of new technology,” and that the agency would still need an organization like ORD. According to Gene Poteat, David did a “great disservice by wiping out the only organization of R&D people who knew the intelligence business.”
⁷⁰

Both of David’s immediate predecessors also questioned her decision to abolish ORD. Evan Hineman called it a “great mistake,” but one that would not be noticed for some time. He noted that there was some discussion in the late 1970s and early 1980s about doing away with the office, but the argument that the other offices conduct R&D activities was a false argument. He expressed concern that “today’s” activities would take budgetary precedence—which David hoped to prevent through the IPO—and noted that during his tenure, 2.5 percent of the agency’s budget was devoted to R&D, but rather than the figure reaching 5 percent as he envisioned, the practice of reserving a specific percentage of the budget for ORD fell victim to budgetary pressure and was ultimately done away with.
⁷¹

James Hirsch also believed the decision was a mistake. He argued that technical experts needed to talk with other technical experts and maintain a certain distance from their customers (the individual offices). He also noted that ORD would usually spend twice its allotted budget, because it received funds from other agencies to manage projects.
⁷²

The abolition of ORD had a ripple effect, leading to the disestablishment of the Office of Advanced Projects shortly before David left office. According to David, the advanced projects office was no longer needed as a bridge between ORD and its consumers.
⁷³

In late June 1998, a CIA press release announced that Ruth David’s tenure as Deputy Director for Science and Technology would end that September. She would be departing to become the President and Chief Executive Officer of ANSER, a nonprofit research institute established in 1958 to conduct studies for the Air Force; it subsequently added the Defense Department and other federal agencies to its list of clients.
¹

The press release contained praise from David’s boss, DCI George Tenet, who expressed his gratitude for “the wise counsel she has given our intelligence collectors and analysts.” He explained that “Dr. David came to the Agency at a time when we needed a leader who could guide the DS&T through major geopolitical transformations that are profoundly affecting how we conduct our mission.” The DCI credited her with developing and delivering “the capabilities our collectors and analysts need to do their critical work in this new and fast-changing environment.”
²

Tenet may have been pleased with David’s accomplishments, but many in the directorate were not. The controversy over FBIS, the loss of NPIC, and the closure of ORD could not but help hurt morale in at least some segments of the directorate. Many veterans undoubtedly would echo the question of one retired directorate official, who wondered, “How can you have any morale if you keep giving everything away?”
³

Of course, David came into office with a restructured NRO a fait accompli, and she did not give away NPIC—John Deutch did. In addition, her outsider status, and possibly her gender, made it even more difficult for her to lead the directorate. Perhaps a longtime, well-respected directorate veteran might have made what has been not only a transformation but a decline in its status easier for the rank and file to accept.

It can also be argued that increased attention to information technology, as recommended by one of Jim Hirsch’s blue-ribbon panels, should be pursued aggressively in order to cope with the information revolution that
has resulted in the availability of an overwhelming volume of data. If the increased emphasis on information technology provides innovative solutions to that problem, it could prove to be at least one significant accomplishment on David’s part.

At the same time, there may be merit to the complaints that she did not appreciate the CIA’s culture or the DS&T’s history. David’s decisions with regard to FBIS and ORD clearly rubbed salt in existing wounds. In the view of one former ORD director, she failed to appreciate that there was more to the directorate than “getting into people’s computers”—that the directorate involved dozens of specialties.
⁴

However much Tenet may have valued David’s leadership, the DCI apparently did not consider finding a replacement to be a matter of great urgency. The directorate was left in the hands of an acting deputy director—Joanne Isham, who had served as David’s deputy since February 1996, initially focusing on resource management issues. Isham, with a B.A. in government and international studies from Notre Dame, had begun her career at the CIA in 1977. Her assignments included several in the Office of Development and Engineering—as chief of security for the Collection Systems Group, as a reconnaissance program manager, and with the Data Communications Group. She also served as the first head of NRO’s public affairs office as well as head of its legislative affairs office. In 1993 and 1994, she was deputy director of resource management for the Community Management Staff. Prior to rejoining the S&T directorate, she served as the CIA’s Director of Congressional Affairs.
⁵

That Tenet did not consider the DS&T as critical a component as it once was possibly explains why the strategic plan that he introduced in spring 1998 focused on the operations and intelligence directorates. At one point, as the days and weeks and then months dragged on without David’s replacement being chosen, Evan Hineman suggested to Tenet that he should either find a new head for the directorate or abolish it.
⁶

Tenet chose the first alternative. In late March 1999, the CIA announced that Dr. Gary L. Smith, the recently retired director of the Johns Hopkins University Applied Physics Laboratory (APL), would join the agency later that spring as the new deputy director for science and technology. Smith, who received bachelor’s, master’s, and doctorate degrees from the University of California at Davis, joined APL in 1970, where he
initially was involved in theoretical and experimental research on detecting submerged submarines.
⁷

Smith was coming from a position that seemed to provide perfect training for leadership of the science and technology directorate. APL has a staff of about 2,700, over 60 percent of whom are engineers and scientists. The laboratory describes itself as “a technical resource to the Department of Defense . . . for innovative research and development.” It designed, for example, the Midcourse Space Experiment (MSX) satellite, which was orbited in April 1996. The MSX carried three sensors whose primary mission was to collect data on ballistic missile signatures.
⁸

But Smith’s particular background and the CIA press release concerning his appointment suggested that developing collection systems would not be Smith’s primary focus. The release noted that in his tenure at APL, Smith had forged “strong ties with the national security community, including operating forces of the military, with senior decisionmakers in a broad range of government agencies, and with Congress.” Even more to the point was the praise for Smith’s building of “new relationships with commercial and industrial research sponsors” and his role in “commercializing [the lab’s] technology breakthroughs.”
⁹

In the release, Tenet expressed his confidence that Smith would “lead the Directorate of Science and Technology into the next century with equal foresight, boldness and agility” as he displayed at APL. Tenet added that he had no doubt that under Smith’s leadership, the directorate “will carry on its proud tradition of putting technology to work enhancing the effectiveness of clandestine collection and all-source analysis.”
¹⁰

But Smith would not be leading the directorate into the next century with boldness or any other quality. On January 12, 2000, DCI George Tenet announced that Smith, after only nine months in office, had turned in his resignation. The DCI explained that Smith would “like to continue his interrupted retirement.” Tenet told CIA employees that “Gary has facilitated an effective transition . . . I know you all join me in thanking Gary for his contributions.”
¹¹

Smith’s sudden departure raised eyebrows both within and outside the CIA, as some wondered if there was something more to the abrupt resignation than a desire to return to retirement—Smith left without either cleaning out his office or submitting a letter of resignation. The story that circulated among a number of present and former CIA personnel is that Smith was never welcomed into the agency’s inner circle, and was
abruptly fired by DCI George Tenet for pushing him in directions he did not want to go.
¹²
*

This time there was no delay in finding a new deputy director. Along with his announcement of Smith’s departure, Tenet informed the CIA’s employees that Joanne Isham would replace Smith—as deputy director, not acting deputy director. Of Isham, he said, “I have no doubt that under Joanne’s leadership, the DS&T will carry on its great tradition of putting technology to work to enhance the effectiveness of clandestine collection and all-source analysis—the key components of our Strategic Direction efforts here at CIA.”
¹³
Significantly, he made no reference to directorate efforts with regard to the development of collection systems or the collection of intelligence—which in the past were the key components of the directorate’s activities.

In explaining Isham’s selection, a CIA spokesman noted that Tenet was comfortable with her, that she was well-respected, and that she had been “outstanding” as David’s deputy. In addition she had both CIA and community experience.
¹⁴
Mark Lowenthal, former State Department intelligence official and House intelligence committee staff director, praised the appointment, observing that “she understands collection . . . what the systems are designed to do, and . . . the use to which the intelligence is being put.” However, Gordon Oehler, former head of the CIA’s Non-Proliferation Center, characterized Tenet as not understanding the importance of science and technology and his appointment of Isham, a nonscientist, as sending the message that he was “no longer interested in S&T.” Today, the directorate, Oehler laments, is a “mere shadow of itself.”
¹⁵

With Isham moving up to the top job, James Runyan, the director of the Office of Technical Collection, became her associate deputy director. Run-yan joined the CIA in January 1997 as deputy director of OTC and became its director in September 1997. Before joining the agency, he had spent thirty years with the National Security Agency, during which time he “developed and deployed collection systems, and managed NSA efforts concerned with field, remote, and special collection responsibilities.”
¹⁶

The bureaucratic loser was Dennis Fitzgerald, who served simultaneously as the head of OD&E and the NRO’s SIGINT directorate, and whose
operation of collection systems. Fitzgerald had been passed over for the top DS&T job twice in a short period of time—first, in favor of Smith, and then in favor of Isham.
¹⁷

By the end of September 2000 when one former senior DS&T official spoke to some current employees of the directorate, they expressed the fear that abolition of the directorate might be no more than two weeks away.

It is not clear that abolition was ever a serious possibility. In any case, two weeks later, a letter from Isham to former senior S&T leaders noted that she and her deputy had been assessing the state of the DS&T and that “it has become clear to us that some changes are necessary if we are to operate successfully in tomorrow’s dynamic environment.”
¹⁸

Those changes focus on achieving “five essential goals” and are to be implemented through some significant organizational changes. The goals include establishing “a single point of entry” into the directorate for new requirements and a central hub for monitoring DS&T responsiveness, combining complementary activities (so as to increase communication and collaboration across organizational lines), integrating information technology activities, revitalizing research and development, and “develop[ ing] the work force of the future.”
¹⁹

Achieving those goals are the responsibility of several new positions and offices. The Program Analysis and Systems Engineering Staff will receive new requirements and monitor the DS&T’s performance. Meanwhile, the Office of Advanced Analytical Tools (AAT) has been supplanted by the Office of Advanced Information Technology (AIT). Unlike the analytical tools office, AIT will not be a joint enterprise with the Directorate of Intelligence. AIT is a combination of AAT, the DS&T Information Services Center, and In-Q-Tel.
²⁰

In February 2001, Larry Fairchild, the director of AIT, noted that the CIA was not growing at a fast rate, but “the amount of information that comes into this place is growing by leaps and bounds.” Among the information tools AIT had under development, Fairchild disclosed, was a computer tool designated “Oasis,” which converted audio signals from television and radio broadcasts into text. Oasis can differentiate accented English to produce more accurate transcripts and distinguish between male and female speakers and among different individuals.
²¹

Another AIT computer tool is FLUENT, which allows a user to conduct computer searches of documents in a number of foreign languages—including Chinese, Korean, Portuguese, Russian, Serbo-Croatian, and Ukrainian. An analyst can enter English words into the search field and receive relevant foreign-language documents in response. The system then translates the document into English and gives the analyst the option of sending it on to a human translator for more precision.
²²

In-Q-Tel is a nonprofit venture-capital firm that the CIA created in late 1999 and gave $28.5 million in agency funds. It is a legacy of Ruth David, who has recalled that the agency was accustomed to working with large defense contractors but not with the newer, smaller firms involved in information technology innovation. In 2000, In-Q-Tel heard from about 500 vendors who proposed projects that they believed would benefit the agency. Twelve development projects were actually funded.
²³

One In-Q-Tel–assisted project involved a commercial search engine named NetOwl, developed by SRA International in Fairfax, Virginia. NetOwl uses natural-language processing in place of keywords to locate information, and can deduce that a word is a name, an organization, or a place. In-Q-Tel funding has permitted SRA to increase the power of NetOwl—allowing it to identify events and relationships and create structured data from unstructured text. According to SRA’s vice-president, Hatte Blejer, it could search the Internet to provide an answer to a question such as “Which high-tech companies were established in northern Virginia last year?”
²⁴

In-Q-Tel funding was also instrumental in allowing companies to develop the Presidential Intelligence Briefing System (PIDS), which is used to produce the President’s Daily Brief (PDB) that only sixteen senior government officials are permitted to read. Previously, CIA analysts had to shuffle hundreds of intelligence cables every day to produce the PDB. The PIDS brings the cables into a Lotus Notes database, performs a variety of search and analysis functions, and then places the brief on a notebook computer. According to Lou Clark, a program manager at In-Q-Tel, PIDS delivers briefs that are more timely because information can easily be added up to the last minute.
²⁵