Preempting terrorist attacks means taking measures to stop a planned attack from being carried out. While prevention as a strategy to combat terrorism takes place, if at all, in the long term and aims at the conditions that promote terrorism, preemption is short term and aims to stop a specific terrorist act. Preemption is often taken to mean the use of lethal force or even to be a euphemism for assassination. While preempting a terrorist attack might require lethal force, it could also require nothing more than a warning to another country that a terrorist group was planning to use its territory, such as the warning that the United States gave to East Germany in 1986 about the attacks planned by Palestinian terrorists that it supported, or a police raid that arrests suspects before they can carry out a terrorist action.

As we have noted, to the extent that it is possible, prevention presupposes an effective national strategy. Preemption does as well. It is only when we have thought through in a comprehensive manner what we must do to maintain our security that we can accurately weigh the contribution that risky and controversial preemptive acts might make. In addition, more than any other method of combating terrorism, preemption requires specific, accurate, and timely intelligence. While general trend analysis may suffice for prevention, it does not for preemption. A warning to the wrong country or the arrest of the wrong suspects will not stop a terrorist attack. Its reliance on intelligence is the great weakness of preemption. This is particularly true if preemption requires lethal force. We can apologize to another country for a false warning and free mistakenly arrested suspects, but we cannot raise the dead. It is possible to imagine one intelligence report or the coincidence of several presenting overwhelming evidence that an attack was about to take place. But such instances will be rare. Warnings of terrorist attacks are as frequent as they are uncorroborated. For the most part, then, any decision to use lethal force requires balancing the harm that may come to Americans with the harm that we may do to innocents. Because of the opprobrium and political damage that would result from a preemptive strike killing innocents, the United States has, at least tacitly, decided that it is better to risk the harm to Americans than to risk killing innocents. An important consideration for the future, however, is whether terrorist acquisition of weapons of mass destruction will change this balancing of risk and so make preemption using lethal force a more acceptable option. Until then, preemption is likely to mean only warnings to other countries and police raids.

The effort to find an offensive means of dealing with terrorism appears to be suspended, then, between prevention, which is of only limited effectiveness, and preemption, which is of only limited feasibility. Disruption, however, is a middle ground between prevention and preemption. Disrupting terrorist activity means targeting a terrorist organization and taking measures, not to stop one of its particular operations, but to render all its activities more difficult. The ultimate goal is to make the organization ineffective. Unlike prevention, disruption assumes that the terrorist activities have begun and, in fact, that we have been targeted and hit already. Therefore, a strategy of disruption allows us to focus our resources on targeting a specific group. Unlike preemption, disruption does not require that we act before a specific attack takes place and so allows for the gradual build-up of intelligence that permits accurate targeting.

The United States has already pursued at least one effective campaign of disruption against a terrorist group: the campaign in the late 1980s against the ANO, previously described. Intelligence officials involved in this operation believe it was highly effective. One high-ranking intelligence official not directly involved in the operation cited it as one of the most effective counterterrorism efforts that the United States has ever undertaken, noting that it severely hampered ANO operations. In interviews, State Department officials who made the démarches to the foreign governments involved concurred with this judgment.

The ways in which terrorist operations can be disrupted are as diverse as the ways in which terrorists operate and support their operations. In addition to démarches, arrests, and various intelligence operations designed to increase the group's paranoia or weaken its infrastructure, they might include legal actions to keep in port ships that have falsified their logs in order to hide participation in terrorist activities or the sabotage of facilities that support terrorist operations. Such operations might hinder a terrorist group's activities directly or, by sowing discord, indirectly. In addition, such operations could be used to send a signal to governments that support terrorism that we know they are providing this support and can take measures to counter it. The fact that these operations could be clandestine or even covert means the message could be received without the public humiliation that overt military actions or public demands bring. Qadaffi may have felt compelled to sponsor the attack on Pan Am Flight 103 because the raid on Libya was such a stinging public humiliation. If sabotage or other disruptive techniques had been used discreetly, we might have made our point in a way that would not have compelled Qadaffi to respond. Diplomatic representation, buttressed with evidence of the damage done, could have carried the message privately to others who needed to hear it. Such an approach, especially if coupled with carefully targeted economic sanctions, might not even require us to cause much damage to the population of a country that sponsored terrorism. This approach may prove more effective than steering a cruise missile into a nearly deserted office building late at night, which is how the United States responded to the Iraqi assassination plot against President Bush. Technological developments will offer soon, if they do not now, the ability to inflict damage on our enemies in a variety of ways and at minimum risk to U.S. personnel.

Disruption shares two disadvantages with preemption: a dependence on intelligence and a risk of harming innocents. To be effective, a campaign of disruption requires a very good understanding of the organization, personalities, and operations of a terrorist group. Even with a lucky break such as a defector, developing this kind of detailed information is difficult and time-consuming. This suggests that we focus our resources on one or two groups that are particularly dangerous to us. While this may make our disruption efforts more effective, it cannot guarantee that innocents will not be harmed inadvertently, especially if disruption includes sabotage. This is a risk in any use of force, however, which we can only avoid by never using military force when dealing with terrorism or any other threat.