What's needed to fix U.S. counterterrorism intelligence? Not more spies nor the power to assassinate terrorist leaders, say two former CIA officials. Instead, start by giving the director of central intelligence the authority to break down the walls that divide domestic and foreign intelligence gathering.
By John Deutch and Jeffrey H. Smith
The terrorist attacks on the World Trade Center and the Pentagon understandably provoked two reactions-that this was the worst intelligence failure in recent U.S. history and that U.S. intelligence gathering and analysis must be vastly improved. Many proposals have been put forward to improve U.S. intelligence capabilities. In order to sort those that make sense from those that do not, it is important first to understand the constraints the intelligence community has inherited.
The framework for U.S. intelligence was created in a different time to deal with different problems. The National Security Act of 1947, which established the Central Intelligence Agency (CIA), envisioned the enemy to be states such as the Soviet Union and also recognized the importance of protecting citizens' rights [see sidebar on page 67]. The result was organizations and authority based on distinctions of domestic versus foreign threats, law enforcement versus national security concerns, and peacetime versus wartime. The Federal Bureau of Investigation (FBI) was responsible for the former, and the intelligence community-comprising the CIA, the National Security Agency (NSA), the Defense Intelligence Agency (DIA), and other agencies-was responsible for the latter.
Law enforcement's focus is to collect evidence after a crime is committed in order to support prosecution in a court of law. The FBI is reluctant to share with other government agencies the information obtained from its informants for fear of compromising future court action. On the other hand, the CIA collects and analyzes information in order to forewarn the government before an act occurs. The CIA is reluctant to give the FBI information obtained from CIA agents for fear that its sources and methods for gaining that information will be revealed in court.
Clearly, the current structure is ill-suited to deal with catastrophic terrorism. Decisions on intelligence reform will revolve around this question of the proper balance between national security and law enforcement goals. Meanwhile, historical boundaries between organizations remain, stymieing the collection of timely intelligence and warnings of terrorist activity. This fragmented approach to intelligence gathering makes it quite possible that information collected by one U.S. government agency before an overt act of terrorism will not be shared and synthesized in time to avert it.
A word about intelligence "failures" is in order. By the most obvious criterion-the success of Osama bin Laden's operatives on September 11-intelligence and law enforcement failed to protect the public. But only time will tell if the information necessary to predict and stop the attacks was in government hands in advance or reasonably could have been. At some point it will be appropriate to analyze this question. For now, however, such an inquiry would only distract government agents and analysts from the critical task of identifying and preventing future attacks.
GIVING THE CIA THE LEAD
The FBI and CIA have been working to overcome the fragmentation of counterterrorism intelligence efforts through personnel exchanges and joint training. Yet the FBI and the intelligence community still have separate counterterrorism centers. This duplication hardly makes sense. In an era when national security must be the preeminent concern, the director of central intelligence (DCI) should manage a single National Counterterrorism Center that plans intelligence collection for all agencies and produces analysis derived from all sources of intelligence. A committee chaired by the DCI and including the national security advisor, the director of the new Office of Homeland Security, and the attorney general should set the agenda for these activities.
The security services of friendly nations are important sources of information for U.S. intelligence; they know their neighborhoods and have access that U.S. agencies do not. At present, the CIA, NSA, DIA, FBI, and the Drug Enforcement Administration have separate agreements with foreign counterpart organizations to obtain information. These efforts should be coordinated. The DCI's authority and responsibility to plan, monitor, and approve arrangements between all intelligence agencies and their foreign counterparts on all intelligence matters, including counterterrorism matters, should be clarified and strictly enforced.
Judging by their recent articles, some editorial writers apparently believe the collection of intelligence through technical means such as communications intercepts and imagery is not important in the fight against terrorist organizations. In fact, cooperation between human and technical intelligence, especially communications intelligence, makes both stronger. Human sources, or HUMINT, can provide access to valuable signals intelligence, which incorporates primarily voice and data communications intelligence. Communications intercepts can validate information provided by a human source. Any operation undertaken in a hostile environment is made safer if communications surveillance is possible. Currently, the NSA, which is under the authority of the secretary of defense, carries out communications intelligence, and the CIA carries out human intelligence, which is under the authority of the DCI. The secretary of defense and the DCI share authority for setting foreign collection priorities. In the case of foreign threats within the United States, the FBI has primary responsibility for setting collection priorities. Here again, the fragmentation makes no sense when considering the global terrorist threat. The new antiterrorism law took a good first step toward remedying this problem by clarifying the DCI's lead role in setting priorities for wiretaps under the Foreign Intelligence Surveillance Act (FISA) and disseminating the resulting information.
In addition, the Bush administration's current review of intelligence, under the leadership of former National Security Advisor Brent Scowcroft, should recommend greater centralization of intelligence collection and analysis under the DCI. Inevitably, strengthening the authority of the DCI will raise the question of whether this position should be separated from the position of head of the CIA. If the DCI is given budgetary, planning, and management authority over the agencies that are responsible for national-level intelligence, then the positions should be separated, just as the secretary of defense sits above the individual services.
DRAGS ON COVERT ACTION
Fragmentation also impairs covert action-activities the United States undertakes to achieve objectives without attribution. Such action has been associated with past CIA efforts to overthrow, in peacetime, political regimes considered a threat to the United States. The future purpose of covert action will be quite different: to destroy terrorist cells and facilities that may produce or store weapons of mass destruction. The distinction between CIA-sponsored covert action and military special operations will become much less relevant, if it is relevant at all. For larger paramilitary operations, a permanent planning staff under the leadership of the secretary of defense, including CIA and FBI staff members, should be put in place to strengthen counterterrorism covert action.
Current law requires both a presidential finding and reporting to Congress of all CIA covert action. No such rule governs covert military operations. In the fight against terrorists, the CIA and the military will be called to conduct joint covert operations, but the differing approval and reporting requirements of these organizations can hamper cooperation. Congress should consider streamlining the law to remove the artificial distinction.
The September 11 attacks renewed questioning about the adequacy of U.S. human intelligence capability. Use of spies is an essential aspect of combating terrorism, and the intelligence community has neither ignored human intelligence nor neglected to target terrorist groups such as Osama bin Laden's al Qaeda organization. Indeed, there have been notable successes in penetrating terrorist groups and preventing planned terrorist acts, but because they were successes they did not come to the public's attention.
Strengthening human intelligence has been a priority of all DCIs. But human intelligence collection is not a silver bullet that can be separated from other intelligence activities and improved overnight. It takes a long time to build a team of experts who understand the language, culture, politics, society, and economic circumstances surrounding terrorist groups. Furthermore, neither bin Laden nor any other terrorist is likely to confide a full operational plan to a single individual, no matter how carefully placed as a source. Spying requires great skill and discipline, something that cannot be achieved quickly or by throwing money at it. To be sure, the morale of the operations directorate hit an unacceptable low in the early and mid-1990s. But this was not due to reduced budgets or lack of presidential support. The poor morale was due to the discovery within the CIA's ranks of Soviet spy Aldrich Ames in 1994, the revelation of CIA activity in Paris in 1995, frequent investigations by Congress and the CIA's own inspector general, and other events that indicated that professional standards had slipped badly.
HUMINT depends critically on other intelligence efforts. It is generally not decisive by itself, but must be combined with all other sources of information. A prerequisite for good human intelligence is a thorough understanding of the sources of terrorism, and much of this kind of information can be obtained from open sources such as local newspapers in the communities that spawn and protect terrorist organizations. Such analytic information is essential for planning collection strategies, successfully penetrating terrorist groups, and mounting covert operations to disrupt terrorist activities and facilities. Successful human intelligence operations rely critically on intelligence analysis to target their efforts. Thus, rather than creating a separate clandestine service, as some have proposed, the United States should support a stronger, seamless partnership between the CIA's operations and intelligence directorates.
CHANGING THE RULES
The recent terrorist attacks gave new momentum to a debate over three controversial rules governing CIA operations. The first of these governs how CIA case officers in the field may recruit agents. In 1995, the CIA established a policy requiring the Directorate of Operations headquarters to approve the recruitment of sources believed to have serious criminal or abusive human rights records. The officials apply a simple balancing test: Is the potential gain from the information obtained worth the cost that might be associated with doing business with a person who may be a murderer, rapist, or the like? Some believe this rule has constrained case officers from recruiting agents inside terrorist groups and therefore made it harder to predict and preempt terrorist acts, although senior CIA officials maintain that the rules have not reduced the quality or quantity of counterterrorism intelligence. Congress recently considered legislation directing the DCI to revoke the rule, but it ultimately enacted a "sense of the Congress" provision, as part of the new antiterrorism law, encouraging intelligence officers to "make every effort" to "establish relationships" with such individuals.
There are two reasons such rules are necessary. First and most important, case officers have been and will continue to be vulnerable when they enter arrangements with agents who do not necessarily produce valuable or accurate information and later are found to have committed atrocities against U.S. citizens or others. These case officers may be investigated by the CIA inspector general, the Department of Justice, and congressional committees. The overriding purpose of the 1995 recruiting guidelines was to keep case officers from worrying about just this possibility of prosecution. Clearance by the Directorate of Operations protected the case officer in the field. The rules did the opposite of what was feared; they gave case officers the incentive to take risks because approval from Washington meant that headquarters had to stand behind field decisions. It is a sad irony that Congress, while passing one piece of legislation that encourages case officers to take risks in recruiting agents, in another authorized the DCI to pay for personal liability insurance for case officers. Congress seems to be saying, "Go take risks, but if later we don't like the risks you took, you will be investigated. And the government will pay your legal bills." This seems an odd way to motivate case officers in the field.
The second reason for the 1995 rule governing recruiting is efficiency. The CIA should focus on recruiting agents that have access to genuinely important information and reward case officers' efforts for the quality of information collected, not just the quantity. It can be difficult to judge the appropriate balance between recruiting numbers of agents that may be valuable and recruiting a few agents that will be vital. In some cases, one can rely on the judgment of experienced station chiefs. But both prudence and experience suggest that officials at headquarters need to review these judgments.
Another contentious rule has been President Gerald Ford's 1976 executive order barring U.S. intelligence agencies from assassinating foreign political leaders. The horror of the September 11 attacks on civilians prompted many to call for a reversal of this ban to allow assassination of a terrorist leader or a political leader who supports terrorism. This move would be unwise. The United States will win the war on terrorism, but one result of this victory should not be a world in which assassination of political leaders is an acceptable norm of international law-a precedent that could be established by U.S. action.
Moreover, assassination is rarely effective in defeating motivated groups. For example, the murder of bin Laden would not necessarily remove the threat from al Qaeda. However, the executive order does not and should not prohibit targeting individual political or military leaders, including leaders of terrorist organizations, in the process of military operations, which take place during overt hostilities where opposing forces and their political leadership know they are at risk.
A third change in rules concerns wiretaps on foreigners in the United States and U.S. citizens (especially those in U.S. corporations set up as front organizations) who are associated with suspected terrorist groups. In addition to clarifying the DCI's role under FISA, Congress also relaxed the conditions under which courts may authorize warrants for national security wiretaps and searches. The intelligence community must have access to telecommunications and databases so it can track the movements and associations of suspected terrorists operating in the United States. Similarly, corporations such as banks and airlines will increasingly be asked or required to cooperate with authorities to trace suspected terrorists. Vigilance will be required to prevent improper spying on Americans, but it is possible to devise a system to collect large amounts of information without compromising the privacy and rights of American citizens.
UNREASONABLE EXPECTATIONS
A larger question underlying discussions of intelligence reform is, how much should Americans expect from the intelligence community? Over the past two decades, despite organizational handicaps and conflicting authorities, the intelligence community has built up a considerable counterterrorism capability that has resulted in many successes and, as is now apparent, some spectacular failures. Clearly, Congress and the executive branch are ready to grant the intelligence community greater authority to pursue the paramount mission of national security. And there are dedicated, talented men and women who will make every effort to reduce the threat of catastrophic terrorism. But while the American people can be better protected, they should be under no illusion that the intelligence community can remove all risk. Even if we destroy al Qaeda, other terrorist groups could also mount acts of catastrophic terrorism, including attacks on our information infrastructure and the use of biological agents such as anthrax, chemical nerve agents, and perhaps even nuclear weapons.
Fortunately, there are not hundreds of such organizations but perhaps only a few dozen, which makes the intelligence task feasible. But it is unreasonable to expect 100 percent success. Thus, while intelligence is the first line of defense, other counterterrorism efforts are also important, including prevention by deterrence or interdiction, bioweapons defense, and managing the consequences of a catastrophic terrorist attack whenever and wherever it occurs.
John Deutch is a former director of central intelligence (1995-96) and is now institute professor of chemistry at the Massachusetts Institute of Technology. Jeffrey H. Smith is a former CIA general counsel (1995-96) and now a partner in the Washington law firm of Arnold & Porter.