The recent war in Iraq was, among other things, a powerful advertisement for the effectiveness of the United States' storied special operations forces. Americans are just now learning what role these commandos played in the conflict, but already it has emerged that, during the early days of the fighting, they managed to secure crucial airfields in western Iraq, protect the country's oil fields from saboteurs (fewer than 10 Iraqi oil fields were ignited, compared to the more than 700 Kuwaiti fields that were set ablaze in 1991), and, most famously, rescue Private Jessica Lynch from an Iraqi hospital.

Yet these achievements, although impressive, do not fully explain the unprecedented prominence currently enjoyed by special operations forces within the U.S. military. True, such troops may have been well suited to the kind of missions they were given in Iraq. But they also happen to fit precisely into the model of a leaner, more flexible military that Defense Secretary Donald Rumsfeld is fighting to create at the Pentagon. And Rumsfeld has made no secret of his plans to thrust special forces into the lead role in the war on terrorism, by using them for covert operations around the globe.

The special forces' success in Iraq has also obscured a more ominous consequence of their newfound popularity: that expanding their role in the way Rumsfeld intends could be very dangerous for U.S. foreign policy. Thanks to the vagueness of U.S. law governing covert action, using the military for such operations is -- at least under one interpretation of the law -- much easier than using the CIA. And this facility seems to appeal to Rumsfeld. It also means, however, that the Defense Department (at least according to its interpretation of the law) can conduct covert operations abroad without local governments' permission and with little or no congressional oversight or recourse. If Rumsfeld gets his way, administration hawks may soon start using special forces to attack or undermine other regimes on Washington's hit list -- without the sort of crucial public debate that preceded the war in Iraq.


"Covert action" is defined by U.S. law as activity meant "to influence political, economic, or military conditions abroad, where it is intended that the role of the United States Government will not be apparent or acknowledged publicly." Covert actions are thus distinct from clandestine missions: whereas the term "clandestine" refers to the secrecy of the operation itself, "covert" refers to the secrecy of its sponsor; the action itself may or may not be secret. An operation conducted secretly in order to preserve tactical surprise, but then acknowledged by U.S. officials after the fact, would not be considered covert.

Covert operations are thus rooted in the notion of deniability. But deniability, although it can be very useful, is also highly problematic for democracies, since deniable policies by definition lack the kind of accountability democracy requires. Because covert operations are hidden from the public, neither the thinking behind such missions nor their consequences can be publicly debated. Yet the sorts of operations that tend to be conducted covertly -- namely, those that are risky or otherwise problematic -- are the very ones that would benefit the most from open discussion. When they are conducted secretly instead, the results can be embarrassing -- or worse, as with the Bay of Pigs or the capture of a CIA pilot helping rebels in Indonesia in 1958. Deniability can also be used to mask an abuse of government power, such as when the director of central intelligence (DCI), William Casey, and the National Security Council's Lieutenant Colonel Oliver North ran the Iran-contra operation out of the White House in the early 1980s.

In addition, covert action can have dangerous repercussions outside the country. Such operations require sending U.S. agents into foreign countries without the permission of local governments. If the agents' identities are discovered, this can cause enormous policy problems for Washington, whether the local government is an ally or an enemy. The target country could become less willing to cooperate with the United States or could attempt reprisals of a similar nature.

More broadly, covert actions undermine Washington's reputation and credibility in the international community. The Bush administration has denounced other governments for infringing on the sovereignty of third parties: in mid-2002, for example, when Russian attack helicopters and fighter planes chased Chechen rebels over the Georgian border without first getting Tbilisi's permission, a senior U.S. official complained that Moscow "can't just violate [Georgians'] sovereignty. You have to work with them." For the United States to turn around and do the same thing in a different country would set a troubling precedent, undermining Washington's trustworthiness and its supposed commitment to the inviolability of national borders.


The current definition of covert action was adopted as part of the Intelligence Authorization Act for fiscal year 1991, which tried to fill in the gaps in oversight that led to the Iran-contra scandal. The act codified two requirements for any covert action. First, there must be a written presidential finding (which cannot be issued retroactively) stating that the action is important to U.S. national security. Second, the administration must notify the House and Senate intelligence committees of the action as soon as possible after the finding has been issued and before the operation has begun -- unless, that is, "extraordinary circumstances" exist. If they do, the president need only fully inform the committees "in a timely fashion." Although these are not particularly confining restraints (indeed, a report by the Twentieth Century Fund Task Force on Covert Action and American Democracy questioned whether they would even prevent another Iran-contra), they do ensure a certain degree of oversight and accountability.

The 1991 law also expanded the oversight provisions, previously applied only to the CIA, to include "any department, agency, or entity of the United States Government." This move was intended to prevent another scandal like the Iran-contra affair, which had involved military officers and the National Security Council (NSC). Even the expanded law, however, exempts certain actions. Most notably, "traditional ... military activities or routine support to such activities" are deemed not to be covert actions within the meaning of the law and thus do not require a presidential finding and notification of Congress. Although the act itself does not define "traditional military activities," the phrase, according to the conference committee report, was meant to include actions preceding and related to anticipated hostilities that will involve U.S. military forces or where such hostilities are ongoing and where the U.S. role in the overall operation is apparent or acknowledged publicly (whether or not U.S. sponsorship of the individual action is made public).

This language means, first of all, that covert operations conducted by U.S. military forces during wartime do not require a presidential finding or congressional notification. The problem, however, lies in the interpretation of the word "anticipated," since if future military hostilities are anticipated, no finding or notification is required. Actions in support of anticipated fighting are most commonly thought of in the literal sense, as those undertaken to "prepare the battlespace." Indeed, the conference committee report of the 1991 law defines "anticipated" hostilities as those for which operational planning has already been approved. But according to a knowledgeable Pentagon official, some in the Defense Department believe that the act gives them the power to undertake activities "years in advance" of any overt U.S. military involvement.

The Pentagon's interpretation of "anticipated" raises an obvious and important question: in prosecuting the war on terrorism, if special forces conduct an unacknowledged operation in a country where U.S. troops are not already present, who will challenge the Defense Department to prove that the action was taken in anticipation of some future military conflict, and so was not a covert action that requires a presidential finding? Technically, the ultimate arbiters of what constitutes covert action are the House and Senate intelligence committees, which exert a de facto veto through their control of the intelligence budget. The budgets for special forces, however, are controlled by the House and Senate armed services committees. Thus there is a crucial split between who determines whether a military action is covert and who determines whether that action will be funded. And if the different committees disagree, it can cause a congressional deadlock with no obvious solution (although in such situations, the armed services committees usually prevail, since they must sign off on intelligence authorization bills before they go to the floor for a vote).

The Senate intelligence committee recently tried to clarify this gray area by explicitly declaring all unacknowledged special forces activity in foreign countries where U.S. military forces are not already present to be covert action. The Pentagon and the armed services committees, however, strongly disagreed with this new language, arguing that that formulation misconstrued or even ignored the "traditional military activities" exception. And Rumsfeld told a town hall meeting at the Pentagon in August that the Senate Armed Services Committee had assured him that no new restriction on special operations had been enacted.

The declassified portion of the report of the intelligence authorization bill that was finally passed in November 2003 noted only that the intelligence committees reaffirmed the "functional definition of covert action" and that they attached "critical importance to the requirements for covert action approval and notification" laid out in the 1991 act. In its one oblique reference to the oversight debate, the report states, "Neither the Administration nor the Conferees have sought or agreed to modify, amend, or reinterpret the scope of the Act, or approval and notification requirements under the Act." In effect, then, the bill was deliberately crafted so as to avoid the fundamental question of oversight for covert military action.

Despite that result, however, there are indications that the issue is not completely dead. Although committee staff members are extremely reluctant to discuss the issue, a few have confirmed that it has been raised and is being discussed, and some committee members have mentioned the need for additional congressional attention to the issue. In late November, for example, Senator Trent Lott (R-Miss.) told Helen Fessenden of CQ Weekly that he had "concerns that the Pentagon is going its own way in this area," and other members have subsequently confirmed off the record that the committees are looking into the issue. Nonetheless, overcoming the military's objections to greater scrutiny will require considerable effort and political capital. In all likelihood, therefore, the status of covert military operations will remain dangerously ill defined for years to come.


Distinguishing "traditional military activity" from covert action is made even more difficult by the current "war on terrorism." Most administration officials contend that the campaign is exactly that: a war. Therefore, they argue, it clearly meets the "traditional military activities" exception to the covert action definition -- meaning that any actions taken in pursuit of it need not be justified by a presidential finding and Congress need not be informed about them. It is by no means certain, however, whether the war on terrorism should literally be considered a "war" in this context.

In support of the administration's argument, the general counsel's office at the Pentagon has pointed to Senate Joint Resolution 23, which authorized the use of force in responding to the attacks of September 11, 2001. That resolution authorizes the president to use all necessary and appropriate force against those nations, organizations, or persons he determines planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001, or harbored such organizations or persons, in order to prevent any future acts of international terrorism against the United States by such nations, organizations or persons.

Some senators, led by John McCain (R-Ariz.) and John Kerry (D-Mass.), have argued that explicitly tying the resolution to the September 11 attacks kept it from being as broad as the infamous Tonkin Gulf resolution that gave President Lyndon Johnson a free hand to fight the Vietnam War. But legal experts, including Harold Koh, the incoming dean of Yale Law School, contend the opposite: that it grants the president virtually unlimited authority, as long as he "determines" that a particular target has some connection to al Qaeda. And some Pentagon lawyers have interpreted the situation even more broadly, arguing that Bush does not even need the legislation; because of the September 11 attacks, they posit, anything he now does to fight the war on terrorism is part of the self-defense of the United States and, therefore, a "traditional military activity" that does not require a presidential finding.

The problem with this thinking is that the United States has traditionally treated terrorism (when committed by nonstate actors) as a crime, not a military action; indeed, Washington has signed several international agreements based on that premise. Even Bush himself initially referred to the attacks and their perpetrators in criminal terms. Within a few days of September 11, however, he began calling them acts of war. And in his November 2001 Military Order, which provided for the trial by military commission of Taliban and al Qaeda members captured in Afghanistan, he characterized the September 11 attacks as being "on a scale that has created a state of armed conflict that requires the use of the United States Armed Forces."

Some administration critics have argued that the joint resolution passed after September 11 does not meet the criteria of either a declaration of war or of "traditional military activities." According to former CIA Inspector-General L. Britt Snider, a lawyer who, as counsel to the Senate intelligence committee, helped develop the definition of covert action and its exceptions in the 1991 act, the resolution's authorization to use force is too vague. Any covert military operation conducted in the war on terrorism, Snider argues, should therefore constitute a covert action under the law.

One of the problems at the core of this debate is that the legal language defining covert action was written long before the September 11 attacks changed the national security landscape. Many in the administration, on Capitol Hill, and in the general public would no doubt argue that in the fight against terrorism, the U.S. government should be able to use every tool it has without worrying about presidential findings or congressional notification. But when that kind of broad authority is combined with a determination to increase the role of special forces in covert operations and with a general doctrine of aggressive preemption -- as in this administration -- the potential for abuse and the risks to U.S. foreign policy rise to dangerous levels.


Operationally, unconventional warfare is conducted by the military's Special Operations Command (SOCOM), which comprises a number of specialized units such as the Army Rangers and Green Berets, the Navy SEALs, and the Air Force Special Operations Command. In addition, the smaller Joint Special Operations Command (JSOC) specializes in "black" direct-action operations such as hunting terrorists and rescuing hostages. JSOC is made up of three shadowy units, the existence of which the Pentagon does not officially acknowledge: the Army's 1st Special Forces Operational Detachment-Delta (Delta Force), the Naval Special Warfare Development Group (DEVGRU, formerly known as SEAL Team 6), and the Air Force's 24th Special Tactics Squadron. Finally, there is also a highly classified Intelligence Support Activity team (known as ISA, or, more recently, as Gray Fox, although its name is changed so often that it is probably something else by now), which was recently transferred from the intelligence command to SOCOM. It is the JSOC units and Gray Fox that are to play the key role in Rumsfeld's plans for "hunter-killer" teams that will pursue "high-value targets" (terrorists) around the world.

Rumsfeld began building up the military's unconventional forces last year after being incensed by their inability to enter Afghanistan until after the CIA had prepared the ground for them. In the Defense Department's budget request for fiscal year 2004, he asked for $6.7 billion for SOCOM, a 34 percent increase from the 2003 total of $5 billion. Rumsfeld also significantly increased SOCOM's authority by changing it from a "supporting command," which can only contribute to other combatant commands' missions, into a "supported command," which can plan and execute its own independent operations (if authorized by the secretary and, if necessary, the president). This change removed a layer in the chain of command governing special operations forces. The previous chain of command ran from the secretary to a regional unified command (European Command, for instance) and then to SOCOM; now, under the new system, the secretary of defense and SOCOM are directly linked. In giving Rumsfeld more control over special operations, however, the change reduced the number of people reviewing proposed missions. And by cutting out regional commands, it also increased the risk that special forces units will plan missions without properly considering (or being reminded of) possible local repercussions.

Another key part of Rumsfeld's campaign to bolster special forces has been a concerted effort to reinvigorate and strengthen the bureaucracy that governs them. His most notable move has been to replace the commander of SOCOM, General Charles Holland (whom Rumsfeld perceived as too cautious) with Lieutenant General Bryan Brown, who had been the deputy in command of special forces. Rumsfeld also recalled General Peter Schoomaker, a former SOCOM commander, from retirement to become the Army's chief of staff, and made Rear Admiral John Stufflebeam, who had been spokesperson for the Joint Chiefs of Staff, the commander of JSOC. In addition, Rumsfeld installed Thomas O'Connell, a former commander of Gray Fox, as the new principal deputy assistant secretary of defense for special operations and low-intensity conflict (SOLIC). Taken together, these moves reveal Rumsfeld's intention to make special forces more prominent within the military and to make their leadership more proactive -- and more loyal to Rumsfeld himself. Already, SOCOM has responded: in April 2003, Lieutenant General Brown described the command as the "nexus of the Department of Defense's global war on terrorism" and promised that, as the "supported commander" in that war, he will "plan and selectively execute combat missions against terrorists and terrorist organizations around the world."

There seems to be little question that the Pentagon plans to use its expanded special operations forces for what are, effectively, covert operations -- at the expense of the CIA. Indeed, the changes appear to be part of a larger turf war between the two agencies. The centerpiece of the Pentagon's campaign was the recent creation of a new undersecretary of defense for intelligence. This official will ostensibly help integrate the data and analyses produced by the Pentagon's eight existing intelligence agencies, which include the Defense Intelligence Agency, the National Security Agency (NSA), the National Imagery and Mapping Agency (NIMA), and the National Reconnaissance Office (NRO). Although on its face, the move seems logical -- especially since the Department of Defense currently commands roughly 80 percent of the intelligence budget -- the DCI, not the secretary of defense, is, by statute, the official charged with running the entire intelligence community and compiling and disseminating its results. Rumsfeld has argued that he needs someone within the Defense Department to ensure that the military's interests in intelligence collection are made clear, and DCI George Tenet has publicly supported the creation of the new office. Nonetheless, the move has been widely seen as an effort by Rumsfeld to assert greater control over the intelligence community.

Rumsfeld has also fiercely objected to proposed reforms that would weaken his own intelligence portfolio. In late 2001, an intelligence review commission led by former National Security Adviser Brent Scowcroft recommended moving the NSA, the NRO, and NIMA out of the Pentagon and placing them directly under the DCI's control, while also shifting the DCI's responsibility for coordination of the intelligence community to a new National Director of Intelligence (NDI). Rumsfeld was reportedly furious at what he saw as hostile recommendations and has so far managed to keep the Scowcroft commission's recommendations from reaching the president's desk, even though the congressional joint committee investigating the September 11 attacks has also advocated the creation of an NDI. The recent creation of the undersecretary of defense for intelligence, meanwhile, has also been widely interpreted as Rumsfeld's attempt to block the creation of an NDI.


Rumsfeld seems convinced that the military is better suited than the CIA for much intelligence work, especially covert operations. But is he right? One difference between the military and the CIA is in size. The CIA's advocates contend that it is better designed to conduct covert operations because it has less bureaucracy and thus can do things faster, cheaper, and with more flexibility than the military. This was the reason, they contend, that the CIA beat the military into Afghanistan. The CIA also has case officers stationed at U.S. embassies throughout the world who are familiar with local languages and customs and who already have extensive intelligence contacts that the military cannot match. If an American is detained in another country during an operation controlled by (or at least involving) the CIA, the agency can often use its local contacts to secure his or her freedom.

Those arguing for a greater military role respond that the CIA simply does not have sufficient resources to manage the global war on terrorism. The CIA has, at most, 600-700 covert operators, compared to the Pentagon's roughly 10,000 special forces combatants, and the CIA frequently has to borrow special forces personnel to make up its shortfalls.

Covert operations pose special dangers for the military, however, because they often require operating out of uniform. People who join the CIA's operations division accept the risks associated with their work and know that, if they are caught, they will have no international protections. Those who join the military generally do so with a different set of expectations: that if they fight for and defend their country, they will be covered by international protections such as the Geneva Conventions, which govern the conduct of war and the treatment of prisoners. But once soldiers start operating out of uniform, they lose the benefit of those conventions. Although individuals may accept that risk, it nonetheless endangers the protection of all U.S. military personnel serving abroad and could damage military morale as well.

Another danger with using the military instead of the CIA for covert operations arises from the different ways these agencies plan their missions. The CIA's covert activities are developed by an operational planning group, either on its own initiative or in response to a request from above. After passing through several levels of approval within the agency, a plan is then reviewed by the deputies' committee at the NSC (and possibly by the principals themselves) before being passed on to the president. The military, however, does things very differently. Because its primary mission is combat, it has full authority to make its own operational decisions with no outside input or oversight. Covert military operations are thus planned completely within the Pentagon and are approved by the secretary (and the president, if it is deemed not a "traditional military activity" and a presidential finding is required). Although that system may be appropriate for covert operations during wartime, it seems dangerously limited for actions that occur off the battlefield.

For these reasons, both CIA agents and special forces operatives oppose expanding the military's covert role. Despite what Rumsfeld says, they argue that special forces have neither the experience nor the training for covert action. Moreover, if they do undertake the training required, it will diminish their readiness or competence for the highly specialized missions they have traditionally undertaken, such as hostage rescue, close-quarters combat, and dealing with weapons of mass destruction. Unlike with covert operations, however, if the military is distracted from these activities, no other agency will be able to fill in the gap.

Another reason many special forces operatives think the military should not start running its own, independent covert operations (as opposed to joint operations under the CIA's direction, as they do now) is that it would expose those operatives to increased risk. If special forces personnel are detained in another country in the course of an operation the CIA does not know about, the agency might not be able, or willing, to use its contacts to free them. Even worse, if both the CIA and special forces were to operate in the same place but without coordination, they could end up attacking each other as an assumed enemy.

In light of these complications, increasing the covert role of special forces outside of conventional wars raises two fundamental questions that should be publicly debated. First is the question of whether the military should be allowed to encroach on the CIA's mandate. If the answer is yes, Americans should rethink the kind of accountability they want to require for such missions.

Although both the administration and Congress have been reluctant to openly discuss this question, now is the time to revisit it, before the facts on the ground make it moot. Some Pentagon officials argue that it is only right that the intelligence community should have greater restraints on its actions than the military because of past abuses. But that is exactly the point: the legal restrictions exist to prevent more abuses from happening, and merely shifting more responsibility to the military will not guarantee against them -- quite the contrary, since the military currently operates with a freer hand.

Unfortunately, the Pentagon has already begun pushing to further loosen the rules on what it may or may not do. Marshall Billingslea, a former deputy assistant secretary of defense for SOLIC, told the House Armed Services Committee in April 2003, "we ... need to recalibrate our expectations for what we assess as actionable intelligence. We need to lower that threshold to the appropriate level and we need to have a higher threshold tolerance for pain in the event that we miscalculate or things go wrong or people get caught or something happens."

As this testimony suggests, the Defense Department not only wants to conduct more covert operations, but it wants to lower the level of proof required to justify them. This would only raise the risk of ill-conceived operations that damage U.S. interests.

Whether by amending the language regarding covert action or by adopting new language specifically tailored to special forces, Congress should ensure that covert military operations be held to at least the same standard of accountability as the CIA's covert actions. The risks inherent in the types of missions that the Pentagon envisions for its special forces are at least as serious as those arising from the CIA covert operations that prompted the restrictions now in place. The rules, therefore, should be just as strict.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now