The Performance of the Intelligence Community
Members of the commission, with your help, your staff has developed initial findings on the performance of the intelligence community against the danger of Islamic extremist terrorism before the September 11 attacks on the United States. These findings may help frame some of the issues for this hearing and inform your work.
In Staff Statement No. 7 we discussed our initial findings on the work of the CIA as an instrument of national policy, in the areas of clandestine and covert action. Today we focus on intelligence analysis and warning, the collection of intelligence, and the overall management of the intelligence community before September 11, 2001.
This report reflects the results of our work on these issues so far. We remain ready to revise our understanding of these topics as our work continues. This staff statement represents the collective efforts of a number of members of our staff. Kevin Scheid, Lorry Fenner, Gordon Lederman, Lloyd Salvetti and Doug MacEachin did much of the investigative work reflected in this statement.
We built upon the very significant work done on this topic in 2002 by the Congressional Joint Inquiry. All the agencies of the intelligence community made the necessary documents and witnesses available to us, often with a considerable investment of time and effort.
"The Intelligence Community and Terrorism"
Today's intelligence community is a collection of agencies which were largely created to help wage the Cold War. The Central Intelligence Agency was created in 1947. The Department of Defense was created in the same legislation. The signals intelligence agencies were in the armed services and were unified under the National Security Agency in 1952, yet stayed in the Defense Department.
The National Security Act forbids the CIA from performing any internal security functions. Internal security is the province of the FBI. In contrast, the CIA collects foreign intelligence focused on human sources outside of the United States. The CIA is also responsible for analyzing information from all sources to provide objective intelligence for the president and policy-makers.
The Defense Department conducts the vast majority of technical intelligence collection. The National Security Agency, located in that Department, intercepts communications. The recently renamed National Geospatial-Intelligence Agency, also located in DOD, analyzes photographs and other imagery and prepares needed maps. These agencies are supported by the National Reconnaissance Office, which acquires, launches, and manages systems that orbit the earth, and is another DOD agency.
Each of the executive departments involved in national security has its own intelligence agency or bureau. These include the Defense Intelligence Agency, the intelligence branches of the military services, the intelligence divisions of the FBI, and the Bureau of Intelligence and Research within the Department of State.
The United States spends more on intelligence than most nations spend on national security as a whole. Most of this money is spent on intelligence collection, much of it on very expensive hardware, such as systems based in space. Most of the intelligence community's budget is spent in the Department of Defense, in part because the collection systems are mainly managed by agencies set up in that department and in part because of the substantial intelligence organizations created to support the armed forces and the operations of the unified military commands around the world.
The director of Central Intelligence has two sets of responsibilities. First, he leads a particular agency, the CIA. The CIA has special responsibilities for clandestine intelligence collection and covert action, and for independent analysis of foreign developments. Second, as the DCI he has the responsibility to coordinate the efforts of the entire intelligence community, this loose collection of federal agencies and parts of agencies, so that it purposefully and efficiently supports broad national priorities set by the president, the National Security Council, and the Congress. The DCI controls the CIA, but other cabinet secretaries and the FBI director direct their parts of the intelligence community.
One of the intelligence community's priorities in the 1970s and 1980s was the danger posed by international terrorism. In that era terrorism was seen as tied to regional conflicts, mainly in the Middle East, and many of the terrorist groups were sponsored by governments. In 1985, President Reagan created a presidential task force to review U.S. efforts to combat terrorism. The task force was chaired by Vice President Bush. One of its recommendations was to establish an "all source intelligence fusion center for international terrorism."
The CIA then created a Counterterrorist Center, or CTC. In 1989, DCI Webster expanded the scope of the CTC beyond CIA to make it a "DCI Center" with responsibilities for overseeing the intelligence effort across all the intelligence agencies. Employees were detailed to the CTC from across the intelligence community. DCI Webster created additional mission-oriented intelligence community centers such as the Counternarcotics and Counterintelligence Centers. DCI Gates added a Non-Proliferation Center, and other DCIs have devised more centers to try to cope with transnational challenges.
The CTC was organized to combat mainly regional terrorist organizations and the states that supported them. Therefore its focus tended to be traditional, on collection of intelligence in and against particular states. That intelligence would then inform the foreign policy choices about engagement or coercion to influence the behavior of countries like Libya, Lebanon, Syria, Iran and Iraq. As the CIA and the rest of the community struggled to reorient its priorities during the early 1990s, the CTC began to observe a disturbing new trend of shadowy new groups as illustrated by those involved in with Ramzi Yousef in the 1993 attack on the World Trade Center.
"Analysis of a New Danger"
Information comes to intelligence agencies from many sources. These sources include the reports from other U.S. government agencies such as the State Department, from counterparts in foreign security agencies, from human agents, from signals intelligence such as communications, from imagery and from open sources like foreign newspapers. The CIA was originally created, in large part, to sort through all such sources and offer unbiased assessments to the nation's leaders. In other words, although the CIA became and remains a principal collector and operator in its own right, its first duty was to provide integrated analysis.
Analysis is more than a news report. Tactical analysis studies a particular case involving an individual or group as a guide to specific operations. Strategic analysis looks beyond the particular in order to see patterns, notice gaps, or assemble a larger picture on a wider timeframe to guide the development of national policy.
Budget cuts in the national foreign intelligence program from fiscal years 1990 to 1996, and essentially flat budgets from fiscal years 196 to 200 (except for the so-called "Gingrich Supplemental" of FY 1999) caused significant staffing reductions that constrained the numbers and training of analysts. Analysis was already a relatively minor part of intelligence budgets devoted mainly to collection and operations.
Meanwhile, during the 1990s, the rise of round-the-clock news shows and the Internet reinforced pressure on the diminishing number of intelligence analysts to pass along fresh reports to policy-makers at an ever faster pace, trying to add context or supplement what their policy consumers were receiving from the media. Many officials told us that the demands of providing "current intelligence" and briefings to more and more consumers, both in the executive branch and in Congress, drained scarce resources away from systematic, reflective strategic analysis.
In the late 1990s, weaknesses in all-source and strategic analysis were spotlighted by independent panels critiquing the intelligence community's failure to foresee the India-Pakistan nuclear weapons tests in 1998, and its limited ability to assess the ballistic missile threat to the United States in 1999. The first panel was led by Admiral David Jeremiah, the second by Donald Rumsfeld. Both panels called attention to the dispersal of effort on too many priorities, declining attention to the craft of strategic analysis, budget constraints, sophisticated denial and deception efforts by adversaries, and security rules that prevented adequate sharing of information. We found similar shortcomings with the quality of finished intelligence on transnational terrorism prior to 9-11.
While we now know that al-Qaida was formed in 1988 at the end of the Soviet occupation of Afghanistan, the intelligence community did not describe this organization, at least in documents we have seen, until 1999. As late as 1997, the CTC characterized Osama bin Laden as a financier of terrorism. This was at a time when the intelligence community had recently received a major input of new information revealing that bin Laden headed his own terrorist organization, with its own targeting agenda and operational commanders. This new information also revealed the previously unknown involvement of bin Laden's organization in the 1992 attack on the Yemen hotel quartering U.S. military personnel and the 1993 shootdown of U.S. Army Blackhawk helicopters in Somalia, and quite possibly in the 1995 Riyadh bombing of the American training mission to the Saudi Arabian National Guard. Nor had any analysts worked through answers to questions about links between bin Laden and his associates with the bombing of the World Trade Center in 1993 and the Manila airlines plot of 1994.
The most impressive piece of analysis on the emerging transnational terrorist threat was the 1995 National Intelligence Estimate entitled, The Foreign Terrorist Threat in the United States. It judged at the time that: " 1/8T 3/8he most likely threat of an attack in the United States would be from transient groupings of individuals similar to those drawn together by Ramzi Yousef. Such groupings lack strong organization but rather are loose affiliations."
The National Intelligence Estimate warned of terrorist attacks in the United States over the following two years. It was updated in 1997. As we mentioned in Staff Statement No. 5, by early 1997 the United States had received dramatic new information about the organization of al-Qaida and its efforts to mount catastrophic attacks against the United States. The 1997 update failed to reflect this new information. No comprehensive national estimates were subsequently produced on terrorism prior to the attacks of 9-11.
Thousands of particular reports were circulated. A number of very good analytical papers were distributed on specific topics such as bin Laden's political philosophy, his command of a global network, analysis of information from terrorists captured in Jordan in December 1999, al-Qaida's operation style, and on the evolving goals of the international extremist movement. Hundreds of articles for morning briefings were prepared for the highest officials in the government with titles such as "Bin Laden Threatening to Attack U.S. Aircraft 1/8with anti-aircraft missiles 3/8" (June 1998); "UBL Plans for Reprisals Against U.S. Targets, Possibly in U.S.," (September 1998); "Strains Surface Between Taliban and Bin Laden" (January 1999); "Terrorist Threat to U.S. Interests in Caucasus" (June 1999); "Bin Laden to Exploit Looser Security During Holidays" (December 1999); "Bin Laden Evading Sanctions" (March 2000); "Bin Laden's Interest in Biological and Radiological Weapons" (February 2001); "Taliban Holding Firm on bin Laden for Now" (March 2001); "Terrorist Groups Said Cooperating on U.S. Hostage Plot" (May 2001); and "Bin Laden Determined to Strike in U.S." (August 2001).
Despite such reports, and a 1999 paper on bin Laden's command structure for al-Qaida, there were no complete authoritative portraits of his strategy and the extent of his organizations involvement in past terrorist attacks. Nor had the community provided an authoritative depiction of his organization's relationships with other governments, or the scale of the threat his organization posed to the United States.
A few analysts within the CTC were dedicated to working on bin Laden. One of them had developed a lengthy, comprehensive paper on his organization by 1998. Her supervisor did not consider the paper publishable and broke the topic down into four papers assigned to four other available analysts. As an indicator of the scarcity of analysts and the press of current intelligence reporting work, it took more than two years for two of these papers to be published at all. The other two were not finished until after 9-11.
Some officials, including Deputy DCI John McLaughlin, are skeptical about the importance of comprehensive estimates. McLaughlin had been in charge of the estimate process. He told us such estimates are time-consuming to prepare. Judgments are watered down in negotiations. Conclusions may duplicate those already circulated in more specific papers. He and others said that key policy-makers understood the threat.
Other officials, however, stress the importance of such estimates as a process that surfaces and clarifies disagreements. Through coordination and vetting views, the community comes to a collective understanding of the nature of the threat it faces, what is known, unknown, and a discussion of how to close these gaps.
Most important, our interviews of senior policy-makers in both administrations revealed a fundamental uncertainty about how to regard the threat posed by bin Laden and al-Qaida. After 9-11, the catastrophic character of the threat seems obvious. It is hard now to recapture the old conventional wisdom before 9-11. For example, a New York Times investigation in April 1999 sought to debunk claims that bin Laden was a terrorist leader, with the headline: "U.S. Hard Put to Find Proof bin Laden Directed Attacks." The head of analysis at the CTC until 1999 regarded the bin Laden danger as still in the realm of past experience, discounting the alarms about a catastrophic threat as relating only to the danger of chemical, biological or nuclear attack which he downplayed, referring in 2001 before 9-11 to "overheated rhetoric" on the subject.
In other words, before the attack we found uncertainty among senior officials about whether this was just a new and especially venomous version of the ordinary terrorist threat America had lived with for decades, or was radically new, posing a threat beyond any yet experienced. Some pointed out to us that, before 9-11, al-Qaida was considered responsible for the deaths of less than 50 Americans, all of them overseas. Former officials, including an NSC staffer working for Richard Clarke, told us the threat was seen as one that could cause hundreds of casualties, not thousands. Such differences affect calculations about whether or how to go to war. Even officials who acknowledge a vital threat intellectually may not be ready to act upon such beliefs at a great cost or at high risk.
Therefore, the government experts who believed there was such a danger needed a process that could win and acknowledge broad support for their views or at least spotlight the areas of dispute. Such a process could also prompt action across the government. The national estimate process has often played this role, and is sometimes controversial for this very reason. It played no role in judging the threat posed by al-Qaida.
In the CTC, priority was given to tactical analysis to support operations. Although the CTC formally reports to the DCI, the center is effectively embedded in the CIA's Directorate of Operations. The center had difficulty attracting talented analysts from their traditional billets in the agency's Directorate of Intelligence. The CTC also was especially vulnerable to the pressures that placed reporting ahead of research of analysis. Strategic analysis was a luxury the strained cadres of analysts in the center could rarely indulge.
In late 2000, DCI Tenet recognized the deficiency of strategic analysis against al-Qaida. He appointed a senior manager to tackle the problem within the CTC. In March 2001 this manager briefed DCI Tenet on "creating a strategic assessment capability." The CTC established a new strategic assessments branch during July 2001. The decision to add about ten analysts to this effort was seen as a major bureaucratic victory. The CTC labored to find analysts to serve in this office. The new chief of this branch reported for duty on September 10, 2001.
"Warning and the Case of Aircraft as Weapons"
Since the Pearl Harbor attack of 1941, the intelligence community has devoted generations of effort to understanding the problem of warning against surprise attack. Rigorous analytic methods were developed, focused in particular on the Soviet Union. Several leading practitioners within the intelligence community discussed them with us. They have been articulated in many ways, but almost all seem to have at least four elements in common: (1) think about how surprise attacks might be launched; (2) identify telltale indicators connected to the most dangerous possibilities; (3) where feasible, collect intelligence against these indicators; and (4) adopt defenses to deflect the most dangerous possibilities or at least get more warning.
Concern about warning issues arising after the end of the Gulf War led to a major study conducted for DCI Robert Gates in 1992 which recommended several measures, including a stronger National Intelligence Officer for Warning. We were told that these measures languished under Gates' successors. The National Intelligence Officer for Warning yielded responsibility to the CTC in handling warnings related to a terrorist attacks. Those responsibilities were passed to an intelligence community Counterterrorism Board that would issue periodic threat advisories.
With the important exception of analysis of al-Qaida efforts in chemical, biological, radiological and nuclear weapons, we did not find evidence that this process regularly applied the methods to avoid surprise attack that had been so laboriously developed over the years. There was, for example, no evident intelligence community analysis of the danger of boat bombs before the attacks on the USS. Cole in October 2000, although expertise about such means of attack existed within the community, especially at the Office of Naval Intelligence.
Amid the thousands of threat reports, some mentioned aircraft in the years before 9-11. The most prominent hijacking threat report came from a foreign government source in late 1998 and discussed a plan for hijacking a plane in order to gain hostages and bargain for the release of prisoners such as the "Blind Sheikh." As we mentioned yesterday in Staff Statement No. 10, this 1998 report was the source of the allusion to hijacking in the Presidents Daily Brief article provided to President Bush in August 2001.
Other threat reports mentioned the possibility of using an aircraft laden with explosives. Of these the most prominent asserted a possible plot to fly an explosives-laden aircraft into a U.S. city. This report was circulated in September 1998 and originated from a source who walked into an American consulate in East Asia. Neither the sources' reliability nor the information could be corroborated. In addition, an Algerian group hijacked an airliner in 1994 in order to fly it into the Eiffel Tower, but they could not fly the plane. There was also in 1994 the private airplane crashing into the White House south lawn. In early 1995, Abdul Hakim Murad-Ramzi Yousef's accomplice in the Manila airlines bombing plot told Philippine authorities that he and Yousef had discussed flying a plane into CIA headquarters. A 1996 report asserted that Iranians were plotting to hijack a Japanese plane and crash it in Tel Aviv.
These past episodes suggest possibilities. Alone, they are not warnings. But, returning to the four elements mentioned above:
"Intelligence Collection"
The CTC and the larger intelligence community tried to understand the emerging terrorist threat with their traditional collection methods of human source collection, or the use of informants; information provided by foreign intelligence services; signals collection, or the intercept of communications; and, open sources, or the systematic collection of print, broadcast, and, in the late 1990s, Internet information. Imagery intelligence was extremely valuable for targeting cruise missiles, interpreting Predator videos, and identifying training camps in Afghanistan. This form of intelligence collection worked well. But its sustained effectiveness depended on cues provided by other sources of intelligence.
Human source intelligence is conducted by both the CIA and the Defense Intelligence Agency. Gaining access to organizations or individuals who have access to terrorist groups has proven extremely difficult for both the CIA and the DIA. This has led to a heavy reliance on "walk-ins" and foreign intelligence services.
Often, CIA's best sources of information on terrorist organizations have been volunteers or "walk-ins," who approach U.S. personnel at embassies and other places for a variety of reasons. But evaluating these volunteers and walk-ins is a time consuming and sometimes risky proposition. The ratio of valuable information providers to charlatans, fabricators or double agents is about 1 to 10. That is, for every 10 walk-ins only one produces information of value to the intelligence community and U.S. policy-makers. Yet some of the best sources on al-Qaida during the 1990s were walk-ins. One of these individuals, Jamal al Fadl, began providing information in 1996 and has testified in open court.
Foreign security services also play a critical role in understanding the terrorist threat. The United States government relied, and relies, heavily on this assistance. A major function of the intelligence community is the development and maintenance of these information sharing relationships, which may include expenditures to help the foreign agency improve its own capabilities. Before 9-11 the U.S. government developed especially helpful relationships with several governments in the Middle East and Southeast Asia. Where these relationships work, the local services have an enormous advantage in collecting intelligence. Of course, the quality of these relationships varied.
The German government provided the U.S. government information on an individual named "Marwan" who was acquainted with the target of a German investigation. This common first name and a phone number in the United Arab Emirates were provided as a possible lead in 1999. The CTC pursued this lead for a short time but, with the scant information provided, the CTC found nothing to provoke a special effort on this lead. The CIA did not ask any other agency in the intelligence community for assistance. We now know that "Marwan" was Marwan al Shehhi, who later piloted United Airlines flight 175 in to the South Tower of the World Trade Center. He used the UAE telephone number in the period before the 9-11 attacks. We are continuing to investigate this episode.
We also corroborated that some countries did not support U.S. efforts to collect intelligence information on terrorist cells in their countries, or did not share the American assessments of the threat. According to a former chief of the CTC, before 9-11 many liaison services were "highly skeptical" and "frankly thought we were crazy." They saw (Osama bin Laden) as more an "oddball" than a real terrorist threat. This was especially true for some of the European services.
Most importantly, from our interviews it is clear that the community has no comprehensive and integrated foreign liaison strategy. Each agency pursues foreign partnerships unilaterally, and has done so for many years with minimal interagency coordination. Signals intelligence has been another source of terrorist-related information. The United States spends a great deal on signals intelligence capabilities.
Signals intelligence provides global reach through land, air, sea and space-based systems. But U.S. capabilities have been challenged by the use of modern systems and the operational security practiced by the current generation of terrorists. Moreover, serious legal and policy challenges arise for foreign intelligence agencies when dealing with communications between the U.S. and foreign countries. The NSA is also prohibited from collecting intelligence on people residing in the United States, whether they are U.S. citizens or not, without a warrant under the Foreign Intelligence Surveillance Act.
Signals intelligence is a source of measuring "chatter," which is an indicator of terrorist activity. Interpreting chatter is difficult. For example, the press reported that the congressional joint inquiry was told about intercepted communications collected on September 10, 2001, saying "tomorrow is zero hour," and about the imminent beginning of "the match." Additional information later came to light within the intelligence community, however, that suggested this information was connected with the opening of the Taliban and al-Qaida military offensive in Afghanistan against the Northern Alliance, following on the September 9 al-Qaida assassination of the Northern Alliances leader rather than the 9-11 attacks.
Finally, open sources -- the systematic collection of foreign media has always been a bedrock source of information for intelligence. Open sources remain important, including among terrorist groups that use the media and the Internet to communicate leadership guidance. This mission was performed by the Foreign Broadcast Information Service. During the early 1990s, that service had been "shredded," as one official put it to us, by budget cuts. But, by 2001, the FBIS had built a significant translation effort for terrorism-related media. The FBIS believes its charter bars open source collection of foreign language media within the United States.
The management of the intelligence community's collection efforts is critical. Beginning in 1999, both Assistant DCI for Collection Charles Allen and CTC Director Cofer Black devoted significant attention to improving the collection of intelligence against the al-Qaida sanctuary in Afghanistan. In Staff Statement No. 7 we mentioned "The Plan" developed to energize the recruitment of human agents. These efforts complemented ingenuous efforts already under way to improve the collection of signals intelligence. In these SIGINT efforts the CIA relied heavily on its own efforts, sometimes working well with NSA and sometimes quarreling. But they ultimately failed to achieve an adequate combined effort.
There were some commendable initiatives. Backed by the White House, Assistant DCI Allen worked with military officers in the Joint Staff during the spring and summer of 2000 to come up with innovative collection ideas. One of these was the Predator drone that first flew over Afghanistan in September 2000.
Strategic collection management depends upon strategic analysis to define the baseline of what is known, and what is not known, and to guide the setting of clear, agreed requirements. This process did not occur. Assistant DCI Allen concentrated on day-today collection challenges with enormous energy and dedication. However, there was no comprehensive collection strategy to pull together human sources, imagery, signals intelligence, and open sources. Even "The Plan" was essentially a CIA plan, not one for the intelligence community as a whole.
Collection was focused in a way that led analysts to look for the next attack to occur overseas. That was where the CIA and the NSA collected. The FBI, on the other hand, acquired little information on U.S.-based individuals and groups. What was acquired was often not shared with other members of the community, in part because of the reasons discussed in Staff Statement No. 9.
Human intelligence relied heavily on the proxies we discussed in Staff Statement No. 7, on foreign liaison and on walk-ins. The CIA did not send its own officers into Afghanistan to unilaterally collect intelligence themselves. The CIA's capabilities for such direct action overseas were very limited. The military's special operations capabilities to perform such tasks were more formidable, but were not utilized to collect intelligence in Afghanistan.
Signals intelligence collection against terrorism, while significant, did not have sufficient funding within the NSA. The NSA's slow transformation meant it could not keep pace with advances in telecommunications.
One final point on collection. We have devoted most of our attention to organizations in Washington, D.C. Tackling terrorist cells, however, takes place in the field. A few examples illustrate the challenges and dangers of working the terrorist target:
Unfortunately, the problems of coordination and sharing we found in Washington are sometimes replicated overseas as well.
"DCI Tenet's War"
On December 4, 1998, DCI Tenet issued a directive to several CIA officials and the DDCI for Community Management stating: "We are at war. I want no resources or people spared in this effort, either inside CIA or the community." Unfortunately, we found the memorandum had little overall effect on mobilizing the CIA or the intelligence community.
The memo was addressed only to CIA officials and the Deputy DCI for Community Management, Joan Dempsey. She faxed the memo to the heads of the major agencies. Almost all our interviewees had never seen the memo or only learned of it after 9-11. The NSA director at the time, Lieutenant General Kenneth Minihan, told us he believed the memo applied only to CIA and not NSA since no one had informed him of any NSA shortcomings. On the other hand, CIA officials thought the memorandum was intended for the rest of the community given the fact that they were already doing all they could, and they thought that the community needed to pull its weight.
The episode indicates some of the limitations of the DCI's authority over the direction and priorities of the intelligence community. Congress attempted to strengthen his authority in 1996 by creating the positions of Deputy DCI for Community Management and Assistant DCIs for Collection, Analysis and Production, and Administration. Perhaps their authority is not great enough. Perhaps it is not used enough. The vision of central coordination has not been realized.
The DCI did not develop a management strategy for a war against terrorism before 9-11. Such a management strategy would define the capabilities the intelligence community must acquire for such a war from language training to collection systems to analysts. Such a management strategy would necessarily extend beyond the CTC to the components that feed its expertise and support its operations, linked transparently to counterterrorism objectives. It would then detail the proposed expenditures and organizational changes required to acquire and implement these capabilities.
DCI Tenet and the CIA's Deputy Director for Operations told us they did have a management strategy for war on terrorism. It was called "Rebuilding the CIA." They said the CIA as a whole had been badly damaged by prior budget constraints and that capability needed to be restored across the board. Indeed, the CTC had survived the budget cuts with less damage than many other components within the agency. By restoring funding across the CIA, a rising tide would lift all boats. They also stressed the synergy between improvements of every part of the agency and the capabilities that the CTC or stations overseas could draw upon in the war on terror.
As some officials pointed out to us, the trade off of this management approach is that by attempting to rebuild everything, the highest priority efforts might get only an average share, not maximum support. Further, this approach tended to take relatively strong outside support for combating terrorism and tried to channel this support into backing for across-the-board funding increases. Proponents of the counterterrorism agenda might be less inclined to loosen the purse strings than they would have been if offered a convincing counterterrorism budget strategy. The DCI's management strategy was also primarily focused on the CIA.
DCI Tenet and his predecessors had not developed the management and administrative tools to run the intelligence community that most federal departments use the monitor and rationalize their resources against priorities. The intelligence community did not have a financial accounting system, a chief financial officer or a comptroller. The CIA had these tools for its own operations; the intelligence community did not. Instead, to manage the community as a whole, the DCI relied on a variety of financial systems maintained by difference agencies and without standardized definitions for expenditures.
Lacking a management strategy for the war on terrorism or ways to see how funds were being spent across the community, it was difficult for DCI Tenet and his aides to develop an overall intelligence community budget for a war on terrorism.
The administration and the Congress relied on supplemental appropriations to increase counterterrorism funding. While supplementals were a useful one-time plus-up, the DCI was not able to build long-term capabilities.
The community lacked a common information architecture that would help to ensure the integration of counterterrorism data across CIA, NSA, DIA and the FBI, and other agencies. In 1998, DCI Tenet called for such integration in his Strategic Intent for the intelligence community with a vision of great unity and horizontal integration across the community, but the intelligence community did not develop a plan to achieve it before 9-11.
Finally, the community had not institutionalized a process for learning from its successes and failures. We did not find any after-action reviews sponsored by the intelligence community after surprise terrorist attacks such as the Embassy bombings of August 1998 or the USS. Cole attack of October 2000. The community participated in inspector general inquiries conducted by individual agencies, but these reviews were perceived as faultfinding, without enough constructive emphasis on learning lessons and discovering best practices. What we did not find was anything between the extremes of no investigation at all, and an adversarial inquiry triggered by a public outcry. We did not find an institution or culture that provided a safe outlet for admitting errors and improving procedures.
"Conclusion"
Our investigation so far has found the intelligence community struggling to collect on and analyze the phenomena of transnational terrorism through the mid- to late 1990s. While many dedicated officers worked day and night for years to piece together the growing body of evidence on al-Qaida and to understand the threats, in the end it was not enough to gain the advantage before the 9-11 attacks.
"Reforming Law Enforcement, Counterterrorism and Intelligence Collection in the United States"
Members of the commission, with your help, your staff has developed initial findings to present to the public on the FBI's current capacity to detect and prevent terrorist attacks upon the United States. These findings may help frame some of the issues for this hearing and inform the development of your judgments and recommendations.
This report reflects the results of work so far. We remain ready to revise our understanding as our work continues. This staff statement represents the collective effort of a number of members of our staff. Peter Rundlet, Christine Healey, Caroline Barnes, Lance Cole and Michael Jacobson did most of the work reflected in this statement.
We were fortunate in being able to build upon strong investigative work done by the Congressional Joint Inquiry and by the Department of Justice's Office of the Inspector General. We have obtained excellent cooperation from the FBI and the Department of Justice, both in Washington and in six FBI field offices across the United States.
It is important for us to emphasize that during the course of our investigation we met outstanding FBI and Department of Justice employees, including analysts, agents, translators and surveillance specialists, among others who strive daily to overcome great obstacles for little recognition in order to safeguard our country. Their dedication, effort and sacrifice are remarkable.
On September 4, 2001, Robert Mueller became the director of the FBI. Soon after the attacks, Director Mueller began to announce and to implement an ambitious series of reforms aimed at, in his words, "transforming the Bureau into an intelligence agency." The FBI's leadership has set in motion an impressive number of potentially significant reforms. We believe the FBI is a stronger counterterrorism agency than it was before 9/11.
Most of the proposed reforms are a work in progress. Institutional change takes time. In field visits last summer and fall, two years after 9/11, we found there was a gap between the announced reforms at FBI headquarters and the reality in the field. There may have been additional progress since then.
We divide our discussion of these reforms and the FBI's current capacity to detect and prevent terrorist attacks in the United States into the following four broad areas, tracking the critiques in Staff Statement No. 9: management priorities and strategy; intelligence collection and processing; strategic analysis; and knowledge management.
"Management Priorities and Strategy"
After 9/11, the FBI abandoned its former opaque structure of "tiered" priorities in favor of a short, clear list of priorities. It made "protecting the United States from terrorist attack" the number one priority. It downgraded the priority attached to once sacrosanct parts of the Bureau's mission, including general crimes and narcotics enforcement, which are being left more to state and local agencies or the Drug Enforcement Administration.
FBI leadership also moved quickly to centralize the management of the counterterrorism program. This centralization represents a shift away from the pre-9/11 "Office of Origin" model in which the field office that initiated a case maintained control over it. All significant international terrorism cases and operations are directed from FBI headquarters.
Director Mueller explained that "counterterrorism has national and international dimensions that transcend field office territorial borders and require centralized coordination to ensure that the individual pieces of an investigation can be assembled into a coherent picture." Director Mueller has also endeavored to transform the reactive, law enforcement culture of the FBI. In the course of announcing reforms in May 2002, Director Mueller said: "What we need to do better is be predictive. We have to be proactive."
Along with these changes, the FBI has received large increases in funding since 2001. Appropriations to the FBI's National Security program have nearly doubled between September 11 and today. The FBI reports that the number of counterterrorism agents has increased from about 1,350 on 9/11 to nearly 2,400 today. It has also increased the number of analysts and language translators supporting the counterterrorism mission. The FBI has also created a number of specialized counterterrorism units at its headquarters. These include a unit to analyze electronic and telephone communications, a unit to exploit intelligence gleaned from documents or computers seized overseas by intelligence agencies, a surge capacity to augment local field investigative capabilities with specialized personnel, and a section to focus on the financial aspects of terrorism investigations.
Because of Director Mueller's efforts, there is widespread understanding that counterterrorism is the FBI's number one priority. However, many agents in the field were offended by the director's statements that the FBI needs a new, proactive culture. Some agents who had worked counterterrorism cases before 9/11 felt prevention had always been part of their mission.
We also found resistance to running counterterrorism cases out of FBI headquarters. Many field agents felt the supervisory agents in the Counterterrorism Division at headquarters lacked the necessary experience in counterterrorism to guide their work. In addition, because the organizational chart for the Counterterrorism Division has changed many times since 9/11, some field office personnel told us that they no longer have any idea who is their primary point of contact at headquarters.
The expertise of agents, analysts, linguists and surveillance personnel contribute to effective counterterrorism operations. However, FBI personnel continue to be pulled away from counterterrorism to assist on criminal investigations. At present, the FBI attempts to address field office reassignments and disruptions primarily through its inspection process.
Director Mueller believes that, while counterterrorism is the number one priority, all agents should have training and experience in traditional criminal matters. The director expects to implement by October a special agent career track that requires new agents to start at a small FBI office and be exposed to each of the FBI's four program areas for their first three years. The programs are counterterrorism/counterintelligence; cyber; criminal investigative; and intelligence. Thereafter, agents will be transferred to one of the largest field offices with a primary assignment in an area of specialization. The FBI will also require agents who seek to be promoted to Assistant Special Agent in Charge or Section Chief to have an intelligence officer certification.
"Intelligence Collection and Processing"
The FBI is widely regarded as one of the best post-event investigative agencies in the world. Many outside experts spoke to us about the FBI's incredible forensic abilities, as illustrated by the Lockerbie case, which enable agents to piece together evidence of a crime. The question after 9/11 has been whether the FBI can also collect intelligence that will lead to the prevention of attacks.
Director Mueller's articulation of priorities has reached the field. FBI personnel consistently told us the current policy is that no counterterrorism lead will go unaddressed, no matter how minor or far-fetched. They also told us that there should be no backlog on translations for international terrorism cases.
Many agents in the field told us that there is a new aggressiveness in pursuing international terrorism cases and a new push for agents to recruit more sources and assets. Agents are no longer required to open parallel intelligence and criminal cases for each terrorism investigation. The "wall" is down. All international terrorism cases are now treated simply as counterterrorism investigations.
The USA Patriot Act, passed by Congress approximately six weeks after 9/11, provided additional investigative tools and has lowered or removed legal hurdles that were widely believed to have hindered the FBI's intelligence investigations. The Attorney General Guidelines, which set forth the standards and parameters of the FBI's investigative authority, have also been changed by Attorney General John Ashcroft. These Guidelines now allow for greater flexibility in employing investigative methods, such as permitting agents to attend public events and to search the Internet, including publicly available subscription services, before opening an investigation. These legal and policy changes have prompted significant public debate about the appropriate balance of civil liberties, privacy and security. Many of these issues were addressed during the commission's hearing last December.
Nearly all FBI personnel we interviewed praised these legal and policy changes. When pressed to describe which of the new authorities are most helpful to them and how they employ them, however, there was much less certainty. In fact, there appears to be widespread confusion, even among DOJ and FBI personnel, over what the Patriot Act actually allows. Although the FBI has revamped and increased its training programs, the FBI's General Counsel recently conceded that much more training and guidance must be provided to personnel in the field.
Many agents in the field told us that although there is now less hesitancy in seeking approval for electronic surveillance under the Foreign Intelligence Surveillance Act (or FISA), the application process nonetheless continues to be long and slow. Requests for such approvals are overwhelming the ability of the system to process them and to conduct the surveillance. The Department of Justice and FBI are attempting to address bottlenecks in the process.
To develop a collection strategy, FBI headquarters has recently undertaken an Intelligence Capabilities Survey of field office intelligence collection derived from all sources. This survey is an appropriate first step in an effort to obtain a comprehensive view of the FBI's capability to collect intelligence against its investigative priorities and to identify the critical gaps in collection.
Recruitment of sources has increased, but agents recognize more sources are needed. Michael Rolince, who at the time was Acting Assistant Director of the Washington field office, told us that although the FBI knows "ten times" more now about the radical Islamic community in his territory than it did before 9/11, its knowledge is at about 20 on a scale of 1 to 100.
A supervisor of an international terrorism squad told us the FBI has not adequately reached out to the communities in which it should be developing sources. He believes that while agents are complying with the FBI's policy that they investigate any lead that comes in, other systematic collection work is left undone.
There have not been many instances in which the FBI has been able to recruit an asset to go abroad with specific collection requirements. Despite the widespread view that assets and informants are the best source of intelligence on where potential terrorists are and what they are doing, many agents complained to us that the training they received on how to recruit, validate and maintain assets was inadequate.
Another ongoing problem is the shortage of qualified language specialists to translate the intercepts. While highest priority cases are supposed to be translated within 24 hours, the FBI cannot translate all it collects. According to a recent report by the Department of Justice Inspector General, "the FBI shortages of linguists have resulted in thousands of hours of audiotapes and pages of written material not being reviewed or translated in a timely manner." The choice is between forgoing access to potentially relevant conversations and obtaining such conversations that remain untranslated. Despite the recent hire of 653 new linguists, demand exceeds supply. Shortages of translators in languages such as Arabic, Urdu, Farsi and Pashto remain a barrier to the FBI's understanding of the terrorist threat.
In addition, language specialists suffer from not being part of an integrated intelligence program. During our field visits, language specialists told us that their summaries and translations are usually not disseminated broadly, not uploaded into a searchable database and not systematically analyzed for intelligence value. The individual case agent has the responsibility for determining whether the information should be disseminated and to whom. Several language specialists expressed concern that neither the case agents nor the analysts coordinate with them sufficiently. As a result, the language specialists often lack the proper context to understand the significance of otherwise innocuous references they hear or read. Moreover, we have learned that if a language specialist mishandles a translation, there are few checks to catch the error.
Finally, at every office we visited, we heard that there were not enough surveillance personnel to cover the requests to conduct live physical surveillance of identified terrorist suspects. Like the language specialists, surveillance personnel are not treated as part of an integrated intelligence program. In most cases, their logs are not searchable electronically and they do not meet regularly with case agents to learn about the targets and the broader investigation.
"Strategic Analysis"
In response to widely recognized shortcomings, an Analysis Branch was created in the Counterterrorism Division soon after 9/11 with the mission of producing strategic assessments of the terrorist threat to the United States. The College of Analytical Studies also was created at the FBI's Quantico training facility to improve the quality of training for new analysts.
On January 30, 2003, Director Mueller announced what FBI leadership has described as the "centerpiece" of its effort to improve intelligence analysis: the establishment of the Executive Assistant Director for Intelligence. Mueller stated that "the directed and purposeful collection and analysis of intelligence has not previously been a primary mission focus of the FBI." The position was created to provide one official with direct authority and responsibility for the FBI's national intelligence program. The many responsibilities assigned to the new Executive Assistant Director fall into four general areas: intelligence collection, analysis, dissemination and intelligence program management.
In April 2003, Director Mueller appointed Maureen Baginski, a former executive of the National Security Agency, to this new position. Under Baginski, the FBI has embarked on a series of proposals designed to integrate intelligence into the FBI's operations. She has directed that each field office create a centralized intelligence component called the Field Intelligence Group. FBI leadership is also striving to professionalize and elevate the status of analysts.
Agents and analysts in the field had heard of these changes. But many were still confused by the pace and number of changes and are uncertain about their titles and roles. We question whether the new intelligence program has enough staff and resources to serve as an engine of reform. It is too early to judge whether the Field Intelligence Groups will develop into the centralized intelligence components they are intended to become.
We are concerned whether the qualifications, status and role of most analysts in the field have changed in practice. In the past, analysts were often promoted from secretarial and administrative positions, and they too often served as "catchall" support personnel.
We spoke with analysts who were discouraged by the pace of reform. Indeed, we heard from many analysts who complain that they are able to do little actual analysis because they continue to be assigned menial tasks, including covering the phones at the reception desk and emptying the office trash bins. As a consequence, many of the agents have very low expectations about the type of assistance they can get from analysts. Furthermore, there appears to be no process for evaluating and reassigning unqualified analysts. To retain analysts, the FBI will have to provide them with opportunities comparable to those offered by other intelligence agencies.
The FBI reports that its Counterterrorism Analysis Branch at headquarters has produced more than 70 strategic assessments. The demand for tactical analysis and executive-level briefings, however, has made it difficult for senior managers to focus their resources sufficiently on strategic analysis.
"Knowledge Management"
The terrorist attacks of September 11 revealed significant deficiencies in the FBI's information sharing capabilities and processes, both with respect to sharing information internally with FBI components, as well as externally with intelligence and law enforcement partners at the federal, state and local levels. While progress has been made in addressing these deficiencies, problems remain.
Although there are many explanations for the failure to share information internally, one of the most common is the FBI's outdated information technology, the Automated Case Support system in particular. It employs 1980s-era technology that is by all accounts user-unfriendly. More troubling, the system cannot be used to store or transmit top secret or sensitive compartmented information.
For a variety of reasons, significant information collected by the FBI never gets uploaded into the Automated Case Support system, or it gets uploaded long after it is learned. One of the reasons for this is the traditional approach to cases, in which information is treated as "owned" by the case agent and maintained in a paper case file. One official told us that headquarters personnel visiting the field have been amazed at the information they found in the paper files.
Agent after agent told us that the primary way information gets shared is through personal relationships. There does not appear to be any recognition that this system fails in the absence of good personal relationships.
Some steps to address these ongoing problems have been taken. The attempt to centralize control over the field offices has been made, in part, to ensure that all of the counterterrorism information collected is brought together in one place and disseminated. These steps, driven in part by the director's responsibility since 9/11 to brief the President daily on terrorist threats, have helped get information from the field to headquarters.
However, improvements have been slow. Many current officials told us that the FBI still does not know what information is in its files. Furthermore, the Department of Justices Inspector General reported in December 2003 that the FBI had not established adequate policies and procedures for sharing intelligence. The FBI has had a long-standing plan to upgrade its information technology systems.
The FBI has upgraded desktop terminals, established new networks and consolidated databases. However, the replacement of the antiquated Automated Case Support system has been delayed once again. The director recently told us that the new Virtual Case File system, which is supposed to enhance internal FBI information sharing, should be ready by the end of the year.
As we described yesterday, while top-level officials had frequent contacts and exchanges of information, the overall performance of the FBI and other intelligence community agencies in sharing information was troubled. A tradition of protecting information in order to preserve it for trial, concerns about compromising sources and methods, the absence of a reports officer function and the lack of sophisticated information technology systems have all contributed to the FBI's reputation of being what one former NSC official called an information "black hole."
In July 2002, the FBI created the new National Joint Terrorism Task Force at headquarters to "enhance communication, coordination and cooperation between federal, state and local government agencies." At present, this headquarters task force consists of 38 government agencies. Similarly, the FBI has increased the number of Joint Terrorism Task Forces (JTTFs) in the field from 35 before 9/11 to 84 today, with more than 1,500 outside representatives participating on a full-time basis. Although the JTTFs vary in size and focus from office to office, they are designed to be "force multipliers," pooling the expertise from many agencies to assist in the collection and sharing of intelligence related to counterterrorism.
The FBI has also begun to hire and train reports officers. Reports officers glean intelligence from case files, briefing notes and elsewhere; summarize the information; and format it for dissemination to the intelligence and law enforcement communities. Although filling these new positions has gone slowly -- indeed none of the field offices we visited had permanent reports officers in place at the time of our visits -- the program is now under way.
The passage of the USA Patriot Act also has facilitated greater information sharing. The Act provides for the sharing of intelligence information obtained under FISA with FBI criminal agents and Department of Justice prosecutors. The Act also requires the expeditious disclosure of foreign intelligence information acquired during the course of a criminal investigation to the director of Central Intelligence.
Despite all of these efforts, it is clear that gaps in intelligence sharing still exist. Michael Rolince, the acting Assistant Director of the Office of Intelligence, put it more bluntly: We are kidding ourselves if we think that there is seamless integration among all of the agencies. Former Acting FBI Director Thomas Pickard told us that the most difficult thing about information sharing is trying to figure out what information will actually be important to someone else. John Brennan, the director of the Terrorist Threat Integration Center, told us that he is seeing a "cacophony of activities" within the Intelligence Community but no strategy and planning. Coordination and collaboration are insufficient, he told us. A fundamental strategy for joint work, for integration, is key. This is a problem neither the FBI nor the CIA nor any other agency can solve on its own. We found there is no national strategy for sharing information to counter terrorism.
In the field, JTTF members cannot easily obtain needed information from intelligence agencies. They are expected to go through FBI and CIA headquarters. For example, the process of obtaining name traces from the intelligence community is slow and unsatisfactory.
Compounding the problem of inadequate coordination at the field level is the lack of access by field agents to information systems that operate at the top secret level or above. Very few field agents or analysts have access to Intelink, a worldwide web of information classified at the top secret level. Such terminals have to be maintained in Sensitive Compartmented Information Facilities and such spaces in FBI field offices are extremely limited. To get access to such systems, for instance, many agents and analysts on the New York JTTF have to leave their building, cross the street and enter a separate building. They must then go to the secure room, which is barely large enough to accommodate a few people comfortably. Keep in mind that before 9/11, the New York office was the key FBI office working on international terrorism.
Basic connectivity is still a problem for some FBI field offices. The then-acting Director of the Washington field office told us last August that he still could not e-mail anyone at the Department of Justice from his desk. He said that the Washington field office, which is the second largest field office in the country, still has only one Internet terminal on each floor.
The FBI also needs to be able to coordinate effectively with the hundreds of thousands of state and local law enforcement officers around the country to prevent terrorist attacks. In recognition of the need to work better with state and local law enforcement, Director Mueller announced the creation of the Office of Law Enforcement Coordination in December 2001. The FBI also sends an unclassified weekly Intelligence Bulletin to over 17,000 law enforcement agencies in the United States. The FBI has granted clearances to many police chiefs and other law enforcement officials to increase information sharing.
We spoke with several state and local law enforcement officials who told us that the FBI is doing a much better job sharing threat-related information. However, the Inspector General for the Department of Justice found that the reports "varied as to content and usefulness." We heard complaints that the FBI still needs to share much more operational, case-related information. The NYPD's Deputy Commissioner for Counterterrorism, Michael Sheehan, speculated that one of the reasons for deficiencies in this information sharing may be that the FBI does not always recognize what information might be important to others. For example, if a source mentions something about New York City to an FBI agent in Seattle, this information could be relevant for the NYPD and should be shared immediately. Los Angeles Police Department officials complained to us that they receive watered-down reports from the FBI. They said FBI agents have been told for years that their careers would be at stake if they improperly shared information. In their view, despite recent progress this attitude will take a long time to change. We have been told that the FBI plans to move toward a "write to release" approach that would allow for more immediate and broader dissemination of intelligence on an unclassified basis.
Central to the effort to coordinate with state and local officials is the expansion of the JTTFs that now exist in every field office. Indeed, Larry Mefford, the FBI's former Executive Assistant Director for Counterterrorism and Counterintelligence, called the JTTF structure "the foundation of the Bureau's information sharing efforts." All of the outside representatives on the JTTFs have top secret security clearances, just as all FBI agents do, and they may pass along certain information to their home agencies on a cleared and need-to-know basis.
We found, however, that the role of agency representatives varies from office to office. Information sharing is often ad hoc and depends upon the personalities involved. Although the representatives bring additional personnel to the FBI, the JTTF structure has not produced the full cooperation between the FBI and state and local law enforcement.
Most outside representatives on these task forces have an understanding of terrorism that is limited to the cases they are working on. Thus, they can not reasonably be expected to be the conduit for all threat and case information that may be important to their home agency.
One state counterterrorism official told us only a very small percentage of state and local police officers serve on the JTTFs and that "important information obtained from these national investigations does not reach the officers responsible for patrolling the cities, towns, highways, villages and neighborhoods of our country." We heard this concern from other state and local counterterrorism officials. As a result, several state and local law enforcement agencies have begun to develop their own counterterrorism efforts, separate and apart from the FBI.
"Looking Ahead"
Two-and-a-half years after 9/11, it is clear that the FBI is an institution in transition. We recognize Director Mueller's genuine attempts to transform the FBI into an agency with the capacity to prevent terrorism. He has made progress. Important structural challenges remain to be addressed in order to improve the flow of information and to enhance the FBI's counterterrorism effectiveness. These challenges include:
Enter to win tickets to movies, concerts, festivals,
sporting events and more. This month's featured contest: Medieval Times.
Learn
more about events sponsored by The Baltimore Sun and baltimoresun.com,
including Smart Talk.
This month's featured section: UNISON.
