Need Writing Help 11/18/2023

CHAPTER EIGHT COVERT ACTION

Covert action, along with spying, is a mainstay of popular ideas about intelligence. Like spying, covert action is fraught with myths and misconceptions. Even when understood, it remains one of the most controversial intelligence topics.

Covert action is defined in the National Security Act as “[a]n activity or activities of the United States Government 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.” Consider the irony: An activity that the United States will deny having conducted is defined in U.S. law.

Some intelligence specialists have objected to the phrase “covert action,” believing that the word “covert” emphasizes secrecy over policy. (The British had earlier referred to this activity as special political action—SPA.) The distinction is important, because even though these activities are secret, they are undertaken as one means to advance policy goals. This cannot be stressed enough. Proper covert actions are undertaken because policy makers have determined that they are the best way to achieve a desired end. These operations do not—or should not—proceed on the initiative of the intelligence agencies.

During the Jimmy Carter administration (1977–1981), which exhibited some qualms about force as a foreign policy tool, the innocuous and somewhat comical phrase “special activity” was crafted to replace “covert action.” The administration thus substituted a euphemism with a euphemism. But when the Ronald Reagan administration came into office (1981–1989), with different views on intelligence policy, it continued to use “special activity” in its executive orders governing intelligence. (See the Congressional Research Service report by DeVine and Peters in Further Readings for a discussion of various legal definitions.)

Ultimately, what covert activities are called should not matter that much. What is significant is that in making changes in appellation, the United States reveals a degree of official discomfort with the tool.

The classic rationale behind covert action is that policy makers need a third option (yet another euphemism) between doing nothing (the first option) in a situation in which vital interests may be threatened and sending in military force (the second option), which raises a host of difficult political issues. Not everyone would agree with this rationale, including those who would properly argue that diplomatic activity is more than doing nothing without resorting to force. Also, there are military operations, such as the killing of Osama bin Laden, that are conducted by the military (with strong intelligence community support) but are clandestine—that is, secret but attributable. Such operations begin to blur the line between classic military actions and intelligence covert actions.

As with counterintelligence, two pertinent questions are whether covert action was a product of the cold war and whether it remains relevant today. Under the leadership of Director of Central Intelligence (DCI) Allen Dulles during the Eisenhower administration (1953–1961), covert action became an increasingly attractive option. (See chap. 2.) It had both successes and failures but was seen as a useful tool in a broad-based struggle with the Soviet Union. In the post–cold war period, emphasis turned to terrorists, proliferators, and narcotics traffickers. In some situations when dealing with these transnational issues, covert action can be the preferred means of action. The renewed emphasis on nation-state issues makes this debate less important.

THE DECISION-MAKING PROCESS

Covert action makes sense—and should be undertaken—only when tasked by duly authorized policy makers in pursuit of specific policy goals that cannot be achieved by any other means. Covert action cannot substitute or compensate for a poorly conceived policy. The planning process for covert action must begin with policy makers justifying the policy, defining clearly the national security interests and goals that are at stake, and believing that covert action is a viable means as well as the best means for achieving specified ends.

Maintaining a capability for covert action entails expenses, for the operation itself and for the infrastructure involved in mounting the action. Even though covert actions are not planned and executed overnight, a certain level of preparedness (such as having on hand equipment, transportation, false documents and other support items, and trained personnel, including foreign assets) must exist at all times. The operational support structure—which also includes prearranged meeting places, surveillance agents, letter drops, and technical support—is sometimes referred to as plumbing. Forming and maintaining such a standby capability takes time and costs money. But the key question at this point in the decision-making process is whether the cost—both monetary and political—of carrying out a covert action is justified. Both types of cost become especially important when looking at actions that may last for months or longer.

Alternatives to covert action need to be considered. If overt means of producing a similar outcome are available, they are almost certainly preferable. Using overt means does not preclude either covert action if overt means fail or covert action employed in conjunction with overt means, but the overt means should usually be tried first.

Policy makers and intelligence officials examine at least two levels of risk before approving a covert action. The first is the risk of exposure. William E. Colby, perhaps reflecting on the large-scale investigations of intelligence that dominated his tenure as DCI (1973–1976), said a director should always assume that an operation will become public knowledge at some point. A difference clearly exists between an operation that is exposed while under way or shortly after its conclusion and one that is revealed years later. Nonetheless, even a long-delayed exposure may still prove to be embarrassing or politically costly.

The second risk to be weighed is failure of the operation. Failure of this nature may be costly on several levels: in human lives and as a political crisis for the nation carrying out the operation, as well as for those it may be trying to help. Decision makers must weigh the relative level of risk against the interests that are at stake. An extremely risky operation may still be worth undertaking if the stakes are high enough and no alternatives are available. In other words, the ends may justify the means, or at least the risks. For example, in the 1980s the United States was looking for ways to aid the mujahideen rebels in Afghanistan who were fighting Soviet invaders. One option was to arm the rebels with Stinger antiaircraft missiles, which would counter the successful Soviet use of helicopters. But policy makers were concerned that some Stingers would fall into the wrong hands or be captured by the Soviets. Ultimately, the Reagan administration decided to send the Stingers, which helped alter the course of the war. It also left Stingers in the hands of the mujahideen after their victory, but policy makers deemed that a smaller risk than Soviet victory in Afghanistan. According to press accounts, in 2019, the Central Intelligence Agency (CIA) developed a technology that limits the use of antiaircraft missiles like the Stinger to specific geographic areas by using geofencing (see chap. 5), thus bounding concerns about unauthorized use once the weapons have been given over.

Even though intelligence analysis and operations exist only to serve policy, intelligence officers may be eager to demonstrate their covert action capabilities. Several factors may drive officers to do so: a belief that they can deliver the desired outcome, a bureaucratic imperative to prove their value, and their professional pride in doing this type of work. However, unless the operation is closely tied to agreed-upon policy goals and is supported as a viable option by the policy community, it starts off severely hampered. Covert action planners must therefore closely coordinate their plans and actions with policy offices.

Covert actions are extraordinary steps, something between the states of peace and war. That alone is enough to raise broad ethical questions, although the policy makers’ willingness to maintain a covert action capability indicates some agreement among them on the propriety of its use. The specific details of an operation are likely to raise ethical issues as well. Should assistance be given to foreign political parties facing a close but democratic election against communist parties (e.g., France and Italy in the 1940s)? Should a democratically elected but possibly procommunist government be subverted and overthrown (e.g., Guatemala, 1954)? Should a nation’s economy be disrupted—with attendant suffering for the populace—to overthrow the government (e.g., Cuba, 1960s)? Should a group opposed to a hostile government be armed, with a view toward fomenting an insurgency (e.g., Nicaragua, 1980s)? The issues these questions raise are important not only intrinsically but also because of the risk of exposure. How do covert actions fit with the causes, standards, and principles that the United States supports?

The decision making that went into the operation to kill bin Laden is again instructive, even though it was not strictly a covert action. Once the policy makers were fairly convinced that bin Laden was in the house in Abbottabad, Pakistan, Obama administration officials considered several means to kill him: unmanned aerial vehicle (UAV, or drone) strikes, conventional bombs, or a special operations team. Key factors that weighed in favor of the special operations team were the large number of bombs that would be needed to be certain of killing bin Laden, the likelihood of innocent casualties in a large bombing raid, and the importance of being certain that bin Laden was dead, which would be unlikely if the house was obliterated by bombing. At the same time, policy makers had no more than a 60 to 80 percent certainty that bin Laden was in the house but decided to launch the operation anyway.

It is also important to note the central role played by various lawyers when a covert action is being considered. First, does the action violate either U.S. laws or U.S. treaty obligations? Second, is it something deemed permissible for the United States to be doing? Are there precedents? The officers responsible for the covert action are especially interested in the lawyers’ views because the operators do not want to be put in potential legal jeopardy for conducting an illegal operation. So, despite the bravado usually associated with covert action, there is an inherent legal caution during the planning stage. Again, the planning for the bin Laden raid is instructive. According to press reports, attorneys from the National Security Council (NSC), CIA, Department of Defense (DOD), and Joint Chiefs of Staff (JCS) were involved in reviewing all aspects of the operation, including the permissibility of burying bin Laden at sea. (See the article by Charlie Savage in Further Readings.)

In evaluating proposed covert actions, policy makers should examine analogous past operations. Have they been tried in this same nation or region? What were the results? Are the risk factors different? Has this type of operation been tried elsewhere? Again, with what results? Although these are commonsense questions, they run up against a governmental phenomenon: the inability to use historical examples. Decision makers are so accustomed to concentrating on near-term issues that they tend not to remember accurately past analogous situations in which they have been involved. They move from issue to issue in rapid succession, with little respite and even less reflection. Or, as Richard E. Neustadt and Ernest R. May pointed out in Thinking in Time: The Uses of History for Decision Makers (1988), they learn somewhat false lessons from the past, which are then misapplied to new circumstances.

Congressional reaction to covert actions is a bigger issue for the United States than it is for other democracies. The congressional committees that oversee the intelligence community are an integral part of the process, as providers of funding and as decision makers who need to be apprised of planned operations—“significant anticipated activities.” Although congressional support is important, it is not mandatory. The long lead times required for many operations also mean that they can be put into the budget process in advance so that funds can be allocated for them. Assuming that appropriated funds exist and that there are no specific bars to the covert action in question, Congress must be informed but has no approval role. Again, a limited number of congressional leaders were informed about the bin Laden mission prior to its execution. Congressional notification can be limited to the “Gang of 8” or “Gang of 4.” (See chap. 10.)

Covert action does require formal approval in the executive branch. The president must sign an order approving the operation, based on the president’s finding that covert action is “necessary to support identifiable foreign policy objectives of the United States, and is important to the national security of the United States.” In intelligence parlance, this document is called a presidential finding. Congress and the American public did not know that the president signed off on each operation until Secretary of State Henry A. Kissinger (1973–1977) was forced to reveal as much before a congressional committee in the mid-1970s. Presidential findings are now required by law and must be in writing (except for emergencies, in which case a written record must be kept and a finding produced within forty-eight hours). The president may not delegate this responsibility. Covert actions are therefore somewhat personal in nature for the president, as each president will have different thresholds of willingness to take risks, inflict casualties, and so on.

The finding is transmitted to those responsible for carrying out the operation and to the members of the House and Senate Intelligence Committees or a more limited congressional leadership group in a memo of notification (MON). Often, because of the long time lines involved, the congressional committees will already have learned about the operation via the budget process, which includes a review of the year’s covert action plan. Congress may wish to be briefed on the specifics of the finding and the operation. The briefings are advisory in nature. Other than denying funding during the budget process, Congress has no basis for approving or disapproving an operation, unless specific laws or executive orders ban them—such as the acts passed by Congress in the 1980s limiting aid to the contras in Nicaragua or the executive order banning assassination (discussed later in this chapter).

However, should committee members or the staffers raise serious questions, a prudent covert action briefing team reports that fact to the executive branch. This should be enough to cause the operation to be reviewed. The executive branch may still decide to go ahead, or it may make changes in the operation to respond to congressional concerns. According to press accounts, the George W. Bush administration considered a covert action to support certain parties and candidates in the Iraqi election in 2005 but rescinded the action because of congressional opposition.

The covert action policy system, for all of its rules, remains fragile because of its inherent secrecy. The Iran-contra scandal (1985–1987) underscored some of its weaknesses. A Democratic majority in Congress, opposed to support for the contras in Nicaragua, cut off all funding. President Reagan, in his usual broad manner, urged his NSC staff to help the contras “keep body and soul together.” NSC staffer Lt. Col. Oliver L. North did this by soliciting donations from private individuals and foreign governments, alleging that DCI William J. Casey (1981–1987), who died just as the scandal broke, had approved his actions. North also argued that Congress’s restrictions applied to DOD and intelligence agencies, not to the NSC staff. In a parallel activity, the NSC staff pursued clandestine efforts to improve ties to Iran and free hostages in the Middle East, despite earlier objections to this policy by the secretaries of state (George P. Shultz) and defense (Caspar W. Weinberger). Israel shipped antitank missiles to Iran at the behest of the NSC staff, with the United States replacing them in Israel’s inventory. North also became involved in the Iranian initiative and suggested diverting to the contras the money that Iran had paid for the missiles.

Iran-contra pointed up several problems in the covert action process:

Questionable delegations of authority ordered and managed covert actions (the actions of North on the NSC staff).

Presidential findings were postdated and signed ex post facto (the finding authorizing the sale of missiles to Iran).

Disparate operations were merged (using the Iranian money to fund the contras).

The executive branch failed to keep Congress properly informed (disregarding the laws restricting aid to the contras and not briefing on the finding to sell missiles to Iran).

Debates on the worthiness of the respective policies involved in Iran-contra notwithstanding, NSC staff and other executive branch officials violated the law banning aid to the contras, as well as a host of accepted norms and rules in managing the operations.

The 2013 debate on providing covert arms assistance to the Syrian rebels is also illustrative of many of these policy issues. Top decision makers in Barack Obama’s administration had been divided for several years on whether or not to give assistance to the rebels, deterred, in part, by reluctance to get involved in yet another conflict in the Middle East. However, once the Bashar al-Assad government had apparently used chemical weapons against the rebels, President Obama was forced to act, owing to this crossing of his self-proclaimed “red line” on chemical weapons use. According to press accounts, the CIA had already been providing intelligence assistance to the rebels. Administration lawyers held the view that overt arms support was precluded by international law, which forbids aiding rebels to overthrow an established government. Covert aid circumvented this problem. However, some members of Congress questioned whether this very limited program would make any difference on the battlefield and also raised concerns about U.S.-provided weapons possibly falling into the hands of rebels linked to al Qaeda. The Syrian operation thus became—like Afghanistan and Nicaragua in the 1980s—an overtly debated covert action.

The creation of the position of director of national intelligence (DNI) in 2004 raised new questions for the supervision of covert action. The DNI is now the president’s senior intelligence adviser, which presumably includes covert action, one of the most important types of intelligence activities. Operational responsibility for conducting covert action remains within the CIA. The law states that the director of the CIA (DCIA) reports to the DNI, but it does not specify how extensive this reporting requirement is. The law is clear that the DNI does not have operational control over the CIA. Thus, the DNI needs means to gain insight into covert action capabilities and the status of ongoing operations. This became an issue between DNI Dennis Blair (2009–2010) and DCIA Leon Panetta (2009–2011), which the NSC decided largely in favor of the CIA. The CIA remains in charge of covert action and the right to select chiefs of station. The DNI was given a role in assessing and evaluating covert action when requested by the president or the NSC, which was much less than Blair apparently sought.

THE RANGE OF COVERT ACTIONS

Covert actions encompass many types of activities.

Propaganda is the old political technique of disseminating information that has been created with a specific political outcome in mind. Propaganda can be used to support individuals or groups friendly to one’s own side or to undermine one’s opponents. It can also be used to create false rumors of political unrest, economic shortages, or direct attacks on individuals, to name a few techniques. The creation of “fake news” via websites, blogs, and other social media, as Russia apparently did during the 2016 U.S. presidential election, probably falls into this category, although the World Wide Web makes doing so much easier in terms of both means and deniability. For example, there is growing concern about “deep fakes,” which are digitally altered audio or video clips that are very difficult to detect as having been altered. Artificial intelligence techniques, specifically deep learning algorithms, appear to be the means by which deep fakes are produced and may also provide the means to discover these fakes.

These activities all come under the heading of “disinformation,” which is a synonym for propaganda. The Senate Intelligence Committee undertook a detailed study of Russian efforts to conduct “influence operations,” including the use of disinformation, during the 2016 U.S. presidential election. Among the Russian tactics was the targeting of African Americans, which also fed into the goal of suppressing Democratic voter turnout. Unfortunately, the firms that own and manage the main social media platforms—Meta (Facebook and Instagram), Twitter, and Google (YouTube)—have been slow to admit the ways in which their platforms have been misused and to take corrective measures.

Political activity is a step above propaganda, although they may be used together. Political activity enables an intelligence operation to intervene more directly in the political process of the targeted nation. As with propaganda, political activity can be used to help friends or to impede foes. For example, in the late 1940s the United States supplied scarce newsprint to centrist, anticommunist political parties in Italy and France during closely contested elections. The United States has also funneled money to political parties overseas to help during elections. Alternatively, a state can use political activity more directly against its foes in other nations, such as disrupting rallies or interfering with their publications.

The United States has tended to use economic activity against governments deemed to be hostile. Every political leadership—democratic or totalitarian—worries about the state of its economy because this has the greatest daily effect on the population: the availability of food and commodities, the stability of prices, the relative ease or difficulty with which basic needs can be met. Economic unrest often leads to political unrest. Again, other techniques may be used in conjunction with economic activity, such as propaganda to create false fears about shortages. Or the economic techniques may be more direct, such as attempts to destroy vital crops or to flood a state with counterfeit currency to destroy faith in the monetary system. For years, the United States attacked Cuba’s economy directly as well as indirectly via a trade embargo. Economic unrest was also a key factor in U.S. efforts to undermine the government of Salvador Allende in Chile in the early 1970s. Economic destabilization may be more effective against a more democratic state, as in Chile, than against a dictatorship, as in Cuba, which has fewer qualms about inflicting want or privation on its people and is much less responsive to (or tolerant of) popular protests.

Sabotage, the deliberate destruction of property or facilities, has reentered the covert action tool bag in the past several years. There can be aspects of sabotage in economic activity (destroying crops, industrial facilities, etc.), but the current focus of sabotage relates to efforts to stem proliferation. Sabotage can be a straightforward effort to subvert ongoing activities or can be used to sow distrust between a supplier and a would-be proliferator—for example, by sabotaging shipped parts but allowing them to go through. Sabotage requires some detailed knowledge about an activity and its location, as well as access. There have been several press accounts of efforts to slow down Iran’s nuclear program, including through the introduction of computer programs, such as Stuxnet. (See chap. 12.) According to press reports, a similar effort was made against the North Korean nuclear program, which did not succeed, although efforts to impede North Korea’s missile progress had some success. Thus, sabotage and cyberspace can go hand in hand.

Indeed, it is likely that cyberspace will increasingly be seen as one means of conducting several types of covert actions, including propaganda, political activity, economic activity, and sabotage. For the United States, at least, this raises several issues about legal authorities and oversight. Cyber capabilities reside largely in the National Security Agency (NSA), which is both an intelligence agency and a combat support agency, and in several military components, including Cyber Command. For a cyber activity to be considered a covert action, as opposed to a military action, it would have to be conducted by an intelligence agency. As discussed below, covert actions and military actions are subject to different types of congressional oversight. Gen. Keith Alexander, who served as both the director of NSA (usually called the DIRNSA; 2005–2014) and as the commander of Cyber Command (2010–2014), said that cyberattacks on enemy computer systems should be authorized by the president and secretary of defense and not be undertaken at the discretion of military commanders.

Coups, the overthrow of a government, either directly or through surrogates, are another step up the covert action ladder. (See Figure 8.1.) Again, a coup may be the culmination of many other techniques—propaganda, political activity, or economic unrest. The United States used coups successfully in Iran in 1953 and in Guatemala in 1954 and was involved in undermining the Allende government in Chile, although the coup that brought down his government was indigenous.

Paramilitary operations are the largest, most violent, and most dangerous covert actions, involving the equipping and training of large armed groups for a direct assault on one’s enemies. They do not involve the use of a state’s own military personnel in combatant units, which technically would be an act of war. The United States was successful in this type of operation in Afghanistan in the 1980s but failed abysmally at the Bay of Pigs in 1961. The contra war against the Sandinistas in Nicaragua was neither won nor lost, but the Sandinistas were defeated at the polls when they held a free election in the midst of a deteriorating economy. (They regained power in a subsequent election.)

Some nations have also practiced a higher level of covert military activity—secret participation in combat. For example, Soviet pilots flew combat missions during the Korean War against United Nations (primarily U.S.) aircraft, and Russian “volunteers” have been active against Ukraine. This type of activity raises several issues: military action without an act of war, possible retaliation, and the rights of combatants if captured. The United States has largely eschewed the practice of direct covert military intervention because of such complications, preferring to allow intelligence officers to take part in paramilitary activities with indigenous forces.

Paramilitary operations run the risk that the forces being trained and supported will have agendas of their own and may not conduct themselves according to one’s own rules. Given that many paramilitary operations are either insurrections or civil wars, the chances of higher levels of animosity are always present. (A French diplomat in the 1920s cynically observed that civil wars were bloodier than wars between nations because “it was more fun killing people you know.”) Most recently, press accounts have raised concerns about the conduct of Afghan paramilitary units that had reportedly been trained by the CIA.

Paramilitary operations need to be distinguished from special operations forces. The most fundamental and important distinction is that Special Forces are uniformed military personnel conducting a variety of combat tasks not performed by traditional military arms. The United States has a Special Operations Command (SOCOM). Other such forces are the British Special Air and Special Boat Services (SAS and SBS). Paramilitary operations do not involve the use of one’s own uniformed military personnel as combatants. In the war in Afghanistan (2001–2021), the role of paramilitary personnel appeared to be closer to actual combat than was primarily the case in Nicaragua, but their main role remained training, helping to supply, and offering leadership assistance to indigenous forces. The CIA’s paramilitary forces in Afghanistan were part of the CIA Directorate of Operations Special Activities Division. According to press accounts, CIA paramilitary personnel were the first U.S. forces in Afghanistan, establishing contact with members of the Northern Alliance and preparing them for the offensive against the Taliban.

The war on terrorists has focused attention on a covert activity that does not fall neatly into the customary range of actions—renditions. Renditions are the seizure of individuals wanted by the United States. These individuals are living abroad and are not in countries where the United States either can or wants to use legal means to take them into custody. The operations are called renditions because the individual in question is rendered (that is, formally delivered) to U.S. custody. Renditions predate the war on terrorists, during the Clinton administration, although the scale clearly increased after 9/11. U.S. officials have justified renditions as part of the United States’ right to self-defense against peop

Renditions are controversial for several reasons. First, they are extraterritorial actions. In some instances, the foreign government in whose territory the rendition occurred was aware of the operation and looked the other way, allowing the rendition to proceed but preserving its own plausible deniability. In the case of terrorists, some renditions have been controversial because the United States did not retain custody of the suspects but sent them on to their home nations, most often in the Middle East. Rules about custody, civil rights, and limits on interrogation tend to be different in most of these states, with the effect that some rendered suspects have likely been subjected to harsh treatment if not outright torture. Although the United States has sought pledges from these states about how they would conduct interrogations, U.S. officials cannot be present at all times in these countries. Critics charge that the United States is therefore knowingly complicit in torture. Others argue that the United States cannot hold all suspects, that it is doing as much as it can to prevent torture, and that the importance of breaking up terrorist networks and gleaning information about them requires such use of foreign nations. (See chap. 13.)

Since 2005, in both Italy and Germany, judges issued indictments against U.S. intelligence officers for renditions, one in Milan and one in Macedonia. Although the U.S. government made no official response, CIA officials stated unofficially that any rendition would have been known to the governments. The Italian government denied any such knowledge. However, in 2013 Niccolo Pollari, former head of SISMI—Italian Military Intelligence and Security Service—was sentenced to ten years in prison for his role in the Milan rendition. In 2014, Pollari’s sentence and those of four other Italian intelligence officers were vacated on the grounds that they should not have been prosecuted because the case involved classified information. One person sought by Italy, reportedly a CIA officer, was Sabrina de Sousa, who was arrested in Portugal in 2017 on an Italian extradition warrant. After official U.S. intervention, Italy granted de Sousa a partial pardon. The German case ended after the United States made it clear that it would not cooperate in the case and would not hand over U.S. citizens to be prosecuted. According to press accounts, in March 2007, CIA director Michael Hayden (2006–2009) complained to European diplomats about the inaccurate and negative information being generated in Europe about CIA activities. Hayden also said that fewer than 100 people had been held in secret sites and fewer than half of these had been subjected to more intensive interrogation procedures. Referring to a report by the European Parliament on secret rendition flights, Hayden said that fewer than 100 flights concerned renditions to third countries and that these were undertaken with the knowledge and assistance of the countries involved, a point also made by the European report.

It is unlikely that even the number of flights claimed by Hayden could have been conducted without the knowledge of European countries. It is also possible that the leaders of the governments involved, or their intelligence services, would rather not admit their cooperation and perhaps see judicial proceedings as a way of quieting domestic opinion.

The actual techniques used to interrogate terrorists or suspected terrorists have been controversial for some time. These are sometimes referred to as enhanced interrogation techniques (EITs). In the period immediately following the 9/11 attacks, various executive-branch legal counsels approved a variety of EITs, arguing that these did not constitute torture, which is banned under the United Nations Convention Against Torture (1984), to which the United States is a signatory. The Eighth Amendment to the Constitution bans “cruel and unusual punishment.” The legal opinions also found that the captured terrorists were enemy combatants, not criminal defendants. The most controversial of the EITs was waterboarding, which was banned by President George W. Bush in 2006. This was confirmed by President Obama in 2009, when he also banned some other EITs.

The controversy over EITs resumed when the majority (Democratic) staff of the Senate Intelligence Committee undertook a study of the CIA’s detention and interrogation program in 2009. (For the various controversies surrounding this study, see chap. 10.) Portions of the study were released in December 2014. The staff report found that the EITs, some of which were seen as being overly brutal, were not effective in obtaining useful intelligence and that the CIA made unjustified claims about their effectiveness in order to keep the program going. The report also faulted the overall management and oversight of the program. DCIA John Brennan (2013–2017) rebutted these findings, as did former DCIs and DCIAs Tenet, Porter J. Goss, and Hayden and several other former senior CIA officials, asserting that the EITs provided intelligence that thwarted terrorist attacks.

The involvement of psychologists in designing the EIT program was also controversial. The American Psychological Association (APA) commissioned an external review of its role in the EIT program. Press reports in July 2015 said that the review concluded that APA members had cooperated with the program and had issued “loose” ethical guidelines to support it. Several APA executives resigned as a result. In January 2016, the Defense Department asked the APA to revisit its ban on the use of psychologists in national security investigations, arguing that a code of conduct that allowed participation under certain conditions was more in the national interest than a blanket prohibition. In 2017, two psychologists who helped design the CIA program reached an out-of-court settlement with three former detainees.

EITs came up again during Gina Haspel’s confirmation hearings to be DCIA in 2018. Haspel declined to condemn the use of some of the EITs outright but said that the CIA should not have been part of the rendition and interrogation program and would not do so in the future. In 2022, the Biden administration announced that it would not use testimony obtained during EIT treatment in trials at Guantanamo. This decision is not binding on future administrations.

The use of armed UAVs falls somewhere beyond the usual use of force but not quite into the category of covert action. Such flights are sometimes conducted by intelligence personnel, as opposed to military operators, but they are not entirely covert. That is, there is little question about which country is responsible for the air strikes, even though the United States does not acknowledge each attack. A United Nations report found this to be problematic, suggesting that it could lead to a future in which many states felt free to use UAVs in this manner. (See chap. 13.)

Cyber technology is attractive for many aspects of covert action. Cyber can be used for propaganda, spreading false information via social media. Cyber can also be used as part of political and economic activity and to conduct sabotage against various industries. Finally, given the difficulties in determining attribution in cyberattacks, its use enhances the plausible deniability of operations. There have been various press accounts about U.S. covert actions in cyberspace. In 2018, President Donald Trump is said to have given CIA authority to conduct cyber operations. U.S. Cyber Command has apparently made digital incursions into Russia’s electric grid as a warning and also took steps to block Russian attempts to influence the 2018 elections.

Several press sources report that the United States undertook cyber operations against Iran in 2019 in response to Iran’s attacks on Saudi Arabia’s oil facilities. This highlights one of the concerns about using cyber in covert action, especially for sabotage. There are times when an attacker wants to be somewhat overt, in order to send a message about the consequences of further action. However, the ability to be covert and to avoid attribution, or to be able to deny it, also raises the possibility of an endless series of cyberattacks and counterattacks in an ongoing unacknowledged clandestine war.

ISSUES IN COVERT ACTION

Covert action, both in concept and in practice, raises a host of issues. The most fundamental is whether such a policy option is legitimate. Like most questions of this sort, there is no correct answer. The prevailing opinions can be divided into two schools—idealists and pragmatists. Idealists argue that covert intervention by one state in the internal affairs of another violates acceptable norms of international behavior. They argue that the very concept of a third option is illegitimate. Pragmatists may accept the arguments of the idealists but contend that the self-interest of a state occasionally makes covert action necessary and legitimate. Historical practice over several centuries would tend to favor the pragmatists. Idealists would respond that the historical record does not justify covert intervention. (This debate took a curious turn with passage of the 1998 Iraq Liberation Act, in which Congress and President Bill Clinton agreed to spend $97 million to replace the regime of Saddam Hussein—an overt commitment to interfere in Iraq’s internal affairs.)

In several instances during the nineteenth and twentieth centuries, the United States intervened in other nations, primarily in the Western Hemisphere, but these activities were largely overt and usually military in nature. The United States began to use covert action in the context of the cold war. Did the nature of the Soviet threat make covert action legitimate? Did the use of this option—not only directly against the Soviet Union but also in developing nations that were often the battlegrounds of the cold war—lessen the moral differences between the United States and the Soviet Union? Again, there are broad differences of opinion. To U.S. policy makers during the administrations of Harry S. Truman (1945–1953) and Dwight D. Eisenhower (1953–1961), the Soviet threat was so large and multifaceted that the question of the legitimacy of covert action never arose. In any event, both administrations preferred covert action to the possibility of a general war in Europe or Asia. However, some people believe that the use of the covert option blurred important distinctions between the two nations.

Assuming that covert action is an acceptable option, is it circumscribed by the nature of the state against which it is being carried out? Or does this question become irrelevant if one accepts the legitimacy of covert action? For example, the United States used covert economic destabilization against Fidel Castro’s Cuba and Allende’s Chile. Both were communists, but Castro had seized power after a guerrilla war; Allende never commanded an electoral majority, but he had been elected according to the Chilean constitution. Castro turned Cuba into a hostile Soviet base; Allende showed only disturbing signs of friendliness to Castro and to other Soviet allies. Instead of having a second Soviet satellite in the Western Hemisphere, the United States opted to destabilize Allende in the hope of fomenting a coup against him. Should the fact that Allende had been elected according to Chilean law have been sufficient to preclude covert action by the United States? Or were U.S. national security concerns of sufficient primacy to make the covert option legitimate? This was not the first time the United States had intervened in democratic processes. As noted, the United States gave covert assistance in a variety of forms to centrist parties in Europe in the late 1940s to preclude communist victories.

Central to the U.S. concept of covert action is plausible deniability—that U.S. denials of a role in the events stemming from a covert action appear plausible. The need to mask its participation stems directly from the idea that the action has to be covert. If the situation could be addressed overtly, the role of the United States would not be an issue. DCI Richard Helms (1966–1973) held that plausible deniability was an absolute requirement for a covert action but also conceded that it was becoming an outmoded concept because of the expanded requirements for oversight and notification.

Plausible deniability depends almost entirely on having the origin of the action remain covert. Once that is lost, deniability is barely plausible. Deniability may have been sustainable during the 1950s and 1960s, but this has become more difficult since the revelation that the president signs each finding to order a covert action.

The scale of the activity also matters. For example, in the aftermath of the Bay of Pigs debacle, President John F. Kennedy (1961–1963) sought counsel from his predecessor, President Eisenhower. Kennedy defended his decision not to commit air power to assist the invasion on the grounds of maintaining deniability of a U.S. role. Eisenhower scoffed, asking how—given the scale and nature of the operation—the United States could plausibly deny having taken part.

Plausible deniability also raises concerns about accountability. If one of the premises of covert action policy is the ability to deny a U.S. role, does this also allow officials to avoid responsibility for an operation that is controversial or perhaps even a failure? Or does the fact that the president must sign a finding put the responsibility on him or her?

The main controversy raised by propaganda activities is that of blowback. The CIA is precluded from undertaking any intelligence activities within the United States. However, a story could be planted in a media outlet overseas that will also be reported in the United States. That is blowback. This risk is probably higher today with global twenty-four-hour news agencies and the World Wide Web than it was during the early days of the cold war. Thus, inadvertently, a CIA-planted story that is false can be reported in a U.S. media outlet. In such a case, does the CIA have a responsibility to inform the U.S. media outlet of the true nature of the story? Would doing so compromise the original operation? If such notification should not be given at the time, should it be given afterward? The proliferation of social media has also made propaganda easier to create and spread but much more difficult to counter or to stop. Very few voices are seen as being authoritative any longer, and there is a large amount of skepticism about official pronouncements and explanations. In 2018, DOD transferred $40 million to the State Department’s Global Engagement Center to counter foreign propaganda and disinformation overtly.

Not all covert actions remain covert. One of the key determinants seems to be the scale of the operation. The smaller and more discreet the operation, the easier it is to keep secret. But as operations become larger, especially paramilitary operations, the ability to keep them covert declines rapidly. Two operations undertaken during the Reagan administration—aid to the contras in Nicaragua and to the mujahideen in Afghanistan—illustrate the problem. Should the possibility of public disclosure affect decision makers when they are considering paramilitary operations? Or should disclosure be accepted as a cost of undertaking this type of effort, with the understanding that it is likely to be something less than covert and not plausibly deniable? Ironically, both of these operations were known and openly debated.

Despite the desired separation of intelligence and policy, covert action blurs the distinction in ways that analysis does not. Instead of providing intelligence to assist in the making of decisions, through covert action the intelligence community is being asked to help execute policy. Of necessity, it has a role in determining the scale and scope of an operation, about which it has the greatest knowledge. The intelligence community also has a day-to-day part to play in managing an operation.

The distinction blurs further because the intelligence community has a vested interest in the outcome of a covert action in ways that are vastly different from its interest in the outcome of a policy for which it has provided analysis. Covert action is not just an alternative means of achieving a policy end; it is also a way for the intelligence community to demonstrate its capabilities and value that may be more concrete than analysis.

Thus, covert action makes the policy and intelligence communities closer collaborators, as the separation between them diminishes. Conversely, the intelligence community takes on additional responsibilities in the eyes of the policy community. The intelligence community usually bears a greater burden for a less-than-successful covert action than it does for less-than-perfect intelligence analysis.

Paramilitary operations raise numerous issues. In addition to the problem of keeping them covert and the strains they put on plausible deniability, paramilitary operations raise serious questions about the amount of time available to achieve their stated goals. Unless these operations appear to have a reasonable chance of success in a well-defined period of time, policy makers find their ensuing options limited. On the one hand, they can decide to continue the operation even if the chances of success—usually defined as some sort of military victory—appear slim. It may be that the paramilitary force is unlikely to be defeated but unlikely to win, offering the prospect of an open-ended operation. On the other hand, policy makers can decide to terminate the operation. U.S. abandonment of the Kurds in Iraq in the 1970s is a case in point. The United States had been supporting the Kurds in their struggle against Iraq to create an independent homeland. Covert aid was given to the Kurds via Iran, which also had an interest in weakening its neighbor. However, the Kurdish effort was inconclusive. In the mid-1970s, the shah decided to resolve his differences with Iraq and ordered the operation to cease. The United States complied, abruptly leaving the Kurds to fend for themselves. But when an operation such as this is shut down, extricating all of the combatants may not be possible. In such a case, what is the obligation of the power backing the operation to the combatants? Do the combatants understand the risks they have undertaken, or are they simply assets of the power backing the covert action?

Within the United States, a long-standing debate has taken place about which agency should be responsible for paramilitary operations: the CIA or DOD. The CIA has traditionally run paramilitary operations because, initially, DOD wanted no involvement in them. If covert action is an alternative to military operations, DOD might find it difficult to keep the two options separate. International law poses another difficulty. Although no international acceptance has been given to covert action, the target may consider the use of military personnel (in or out of uniform) in such an activity to be an act of war. Ironically enough, the covert nature of the action also allows the target to ignore or downplay the fact of the intervention, lest relations deteriorate further as they likely would under an overt attack. Finally, the involvement of DOD may undercut the effort to achieve plausible deniability.

However, DOD has greater expertise than the CIA in the conduct of military operations as well as a greater infrastructure to carry them out, which might save money. Removing paramilitary operations from the CIA might spare the intelligence community some internal strains caused by having responsibility for both analysis and operations. New strains might subsequently appear in DOD. The May 2011 attack on bin Laden was a special operations mission, not a paramilitary operation supporting non-U.S. forces in the field.

The war in Afghanistan and the war against terrorists renewed the debate. Secretary of Defense Donald Rumsfeld (2001–2006) pushed for a greater role for SOCOM, including recruiting and maintaining spies in enemy forces. At the same time, the CIA had increased its own paramilitary capability, both as part of DCI Tenet’s overall effort to enhance the Directorate of Operations and to respond to the war on terrorists. In its 2004 report, the 9/11 Commission (National Commission on Terrorist Attacks Upon the United States) recommended that SOCOM take over paramilitary operations from the CIA, based on the view that the two organizations had redundant capabilities and responsibilities. The commission envisaged the CIA organizing paramilitary units while SOCOM was responsible for final planning and execution.

A January 2004 study by the Army War College pointed out some fundamental differences in how the two groups operate, suggesting that even a collaborative effort would be difficult. For example, in jointly conducted operations, would military personnel be covered by the Geneva Convention? Would the necessary secrecy create chain of command problems and make it more difficult to communicate with or to identify friendly units? How would Congress oversee such operations? In February 2005, a study requested by President George W. Bush came out against the recommendations of the 9/11 Commission and argued that the CIA should retain its paramilitary capabilities. In June 2005, the Bush administration confirmed the CIA’s role in covert action. Still, SOCOM can be expected to continue to play a larger part in this area than was the case in the past, probably necessitating some clarification of duties in the future. The war in Afghanistan continued to raise issues of this sort. The House Intelligence Committee, in 2009, said that certain military operations, called Operational Preparation of the Environment (OPE), which appear to be a combination of intelligence collection, covert action, and unconventional warfare, evade intelligence oversight because they are conducted by the military. Press accounts state that the CIA has trained Afghan paramilitary units, and DOD has confirmed that U.S. military personnel have been training Pakistani troops in counterinsurgency warfare. Press accounts have also described a major expansion of military clandestine bases and activities in the Central Command (CENTCOM) and Africa Command (AFRICOM) areas of operations.

Efforts by SOCOM to expand its ability to deploy units overseas more quickly and more independently have raised questions about the relationship of SOCOM and the regional commanders as well as the effect such authority would have on the role of the State Department in decisions about increased overseas deployments. SOCOM’s budget has continued to grow, but some members of Congress have raised concerns about the degree of secrecy in SOCOM budget requests—some of which Congress had sanctioned in the past—which makes congressional oversight of the budget more difficult. A SOCOM program to assess the effectiveness of propaganda was put on hold in 2014 because of congressional concerns.

Other factors of importance in this discussion of special operations versus covert intelligence teams are legal authorities, funding, and congressional notification. The U.S. military operates under Title 10 (Armed Forces) of the U.S. Code; intelligence comes under Title 50 (War and National Defense). The legal authorities for their missions are therefore different and can lead to some difficult, if not comical, debates. A senior former intelligence official described how a cross-border operation involved both special operations and intelligence personnel, so it was not clear if it came under Title 10 or Title 50. The Solomonic decision in this case was as follows: If the operation took place during the day, it would be an intelligence lead and a covert action so as to be able to deny responsibility. However, if it took place at night, it would be a special operations mission, as plausible deniability would be less of an issue. There are also issues of who funds the operation and requirements for congressional notification, which are more stringent for covert action than they are for small-scale military operations—although the May 2011 bin Laden operation was briefed to a small number of congressional leaders because of the risk involved. Recent press accounts suggest an increase in joint intelligence–special forces operations against various terrorist targets.

A related issue has been the relative expertise of the cadre in the CIA responsible for covert action. Like the rest of the intelligence community, this staff, which has always been quite small, suffered losses in the budget cuts of the 1990s. Like the analytic cadre, it underwent fairly rapid growth after the 2001 attacks. President George W. Bush ordered a 50 percent increase in the Directorate of Operations at the time. As with the analysts, this has meant a larger but more inexperienced workforce. This relative inexperience was seen as one of the factors in the attack at a CIA base in Khost, Afghanistan, in December 2009, when a Jordanian agent who had been a source blew himself up and killed seven CIA officers. A review commissioned by DCIA Panetta concluded that the Jordanian had not been sufficiently vetted and that security precautions were insufficient. Several of the post-Khost actions that Panetta announced underscored the need for greater involvement in operations by more veteran officers.

According to press accounts, a CIA study commissioned in 2012 or 2013 found that covertly arming rebels seldom worked, the mujahideen in Afghanistan being a major exception. Efforts by both the CIA and Defense to arm “moderates” among the Syrian rebels fighting the regime of Bashar al-Assad underscore the difficulties of paramilitary operations. The Defense program was abandoned in 2015 as a near total failure, very few Syrians having been trained. The CIA program, another overt covert operation, was hampered by limited funding and by the complexity of the Syrian conflict. Determining which of the many rebel groups was “moderate” was difficult, and each group was so small as to likely make little difference to the overall outcome even if they did receive more arms. The Trump administration ended the Syria program in June 2017, apparently based on the view that it had little success in toppling al-Assad.

As with human intelligence (HUMINT), some covert actions rely on working with foreign liaison services. The issues relating to HUMINT raised earlier remain the same: motivation, reliability, and the ability to keep matters secret, as well as their standards of conduct. In the war on terrorists in Afghanistan and neighboring Pakistan, these issues were of particular concern with Pakistan’s Directorate of Inter-Service Intelligence (ISI), which had been a sponsor and patron of the Taliban prior to 2001. Saudi intelligence appears to have been instrumental in stopping the printer cartridge bombs that were being sent from Yemen as air freight to the United States but were intercepted in Dubai in October 2010. (See chap. 15 for a more complete discussion of foreign intelligence services.)

One concern raised by the conduct of covert actions is their possible effect on intelligence analysis, which is carried out, in part, by the same agency conducting the operation. If the CIA is conducting an operation—particularly a paramilitary operation—is it reasonable to expect analysts of the CIA to produce objective reports on the situation and the progress of the paramilitary operation? Or will there be a certain impetus, perhaps unstated, to be supportive of the operation? DCI Dulles kept the Directorate of Intelligence—the CIA’s analytical arm—ignorant of operations in Indonesia (1957–1958) and at the Bay of Pigs (1961) so as not to contaminate it with knowledge of these operations. This potential problem can be addressed by the creation of mission centers in the CIA, jointly staffed by operations officers and analysts.

In seventeenth- and (to a lesser extent) eighteenth-century Europe, statesmen occasionally used assassination as a foreign policy tool. Heads of state, who were royalty at this time, were exempt from this officially sanctioned act, but their ministers and generals were not. Soviet intelligence occasionally undertook “wet affairs,” as it referred to assassinations. Israeli intelligence has allegedly killed individuals outside of Israel. In 2006, a former KGB officer, Alexander Litvinenko, was assassinated in London via radioactive polonium. A British public inquiry said Litvinenko’s death was a Russian intelligence operation and was probably ordered by President Vladimir Putin. In 2018, another former Russian intelligence officer and spy for Britain, Sergei Skripal, along with his daughter, were hospitalized after exposure to a nerve agent, Novichok. The British government blamed Russia and expelled twenty-three Russian diplomats. One Briton died and another became seriously ill after coming into contact with the bottle apparently used by two Russian agents who had been identified as the ones who carried out the attempt. (See chap. 15 for a fuller discussion of the Russian role.)

The Church Committee (Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities), chaired by Frank Church, D-ID, was formed in 1975 to investigate allegations that the CIA had exceeded its charter. The panel found in 1976 that the United States was involved in or witting of several assassination plots in the 1960s and 1970s—the most famous being that against Fidel Castro—although none succeeded. (See box, “Assassination: The Hitler Argument.”)

Assassination: The Hitler Argument

Adolf Hitler is often cited as a good argument in favor of assassination as an occasional but highly exceptional policy option. But when would a policy maker have made the decision to have him killed? Hitler assumed power legally in 1933. Throughout the 1930s, he was not the only dictator in Europe who repressed civil liberties or arrested and killed large numbers of his own population. Josef Stalin probably killed more Soviet citizens during collectivization and the great purges than the number of people sent by the Nazis to death camps. Deciding to kill Hitler prior to his attacks on the Jews or the onset of World War II would have required a fair amount of foresight as to his ultimate purposes. Little about Hitler was extraordinary until he invaded Poland in 1939 and approved the “final solution” against the Jews in 1942.

Britain revealed in 1998 that its intelligence service considered assassinating Hitler during the war, even as late as 1945. The British abandoned the plan not because of moral qualms or concerns about success but because they decided that Hitler was so erratic as a military commander that he was an asset for the Allies.

Since 1976, the United States has formally banned the use of assassination, either directly by the United States or through a third party. The ban has been written into three successive executive orders, the most recent signed by President Reagan in 1981, and updated in 2008, which remains in effect.

Still, the policy continues to be controversial. Although support for the ban was fairly widespread when instituted by President Gerald R. Ford (1974–1977), debate over the policy has grown. Opponents continue to hold that it is morally wrong for a state to target specific individuals. But proponents have argued that assassination might be the best option in some instances and might be morally acceptable, depending on the nature of the target. Drawing up such guidelines still appears to be so difficult as to preclude a return to the previous policy. In the aftermath of the September 11, 2001, terrorist attacks, debate over the assassination ban was renewed. (See box, “The Assassination Ban: A Modern Interpretation.”) The issue had changed somewhat, however, in that the United States now considered itself to be at war with terrorists, which altered the nature of the target and the legitimacy of using violent force. (See chap. 13 for a more detailed discussion of the ethical and moral issues raised by assassination.)

The Assassination Ban: A Modern Interpretation

In August 1998, the United States launched a cruise missile attack on targets in Afghanistan associated with al Qaeda leader Osama bin Laden. The United States believed that bin Laden was behind the terrorist attacks earlier that month on two U.S. embassies in East Africa.

The Clinton administration later stated that one goal of the raid was to kill bin Laden and his lieutenants. Administration officials also argued that their targeting of bin Laden did not violate the long-standing ban on assassinations. Their view was based on an opinion written by National Security Council lawyers that the United States could legally target terrorist infrastructures and that bin Laden’s main infrastructure was human.

After the September 2001 attacks, bin Laden and other terrorists were seen as legitimate combatant targets, as the United States was at war against them in self-defense, as described by Attorney General Eric Holder after bin Laden’s death in 2011.

On January 3, 2020, an acknowledged U.S. drone strike killed Iranian Gen. Qassem Soleimani in Baghdad. Soleimani commanded Iran’s Quds Force, part of the Islamic Revolutionary Guard Corps (IRGC). (“Al-Quds,” meaning the Holy, is the Arabic and Farsi name for Jerusalem.) The Quds Force conducts clandestine operations outside of Iran. The United States has designated the Quds Force as a foreign terrorist organization since 2007. An official Defense Department statement said that Soleimani was “actively” developing plans to attack U.S. diplomats and service personnel in Iraq, and that he was responsible for the death and injury to numerous U.S. and coalition members to date. The attack “was aimed at deterring future Iranian strike plans.”

To some observers, the killing of Soleimani appeared different from that of bin Laden, who had attacked the United States, and perhaps even from Anwar al-Awlaki (see below), a U.S. citizen who promoted attacks on the United States. Soleimani was a senior foreign official of an admittedly hostile regime, a category that can embrace a large number of individuals. There is little precedent for an operation of this sort—with the exceptions of the active but failed plots against Castro and Patrice Lumumba, then the prime minister of the Democratic Republic of the Congo—including throughout the cold war. According to press accounts, Presidents George W. Bush and Barack Obama had both decided against targeting Soleimani. Also, Congress was not briefed on this operation. It is not clear whether President Trump signed a finding when ordering the attack.

The use of UAVs to kill terrorists is not strictly a covert action, although it does involve acts that the United States does not always acknowledge. However, there is little deniability at present as the United States has been the only nation using armed drones in the areas where these attacks occur—although Britain used drones in September 2015 to kill two Britons fighting for the Islamic State. Prime Minister David Cameron cited self-defense, saying the two Britons had been planning attacks in the United Kingdom. If more states begin using drones unilaterally, there may be a need to reach some sort of international consensus on what types of operations are allowed or not—similar to other rules of war—although this could be difficult to achieve.

According to press accounts, the United States began using armed drones over Pakistan around 2002. The use of armed UAVs continued and increased significantly during the Obama administration. Some strikes targeted specific individuals; others have been “signature strikes” based on observed activities, such as men of military age gathering in groups—the “pattern of life” discussed in chapter 5. The use of armed UAVs became more controversial in 2012–2013 for several reasons. Some—including former DNI Dennis Blair and former Afghan commander Gen. Stanley McChrystal—argued that the number of high-value terrorist targets was decreasing and that the continued heavy reliance on drones was having a negative effect on U.S. relations with Pakistan and other states and was negatively affecting “hearts and minds” in areas where terrorists sought support. Pakistan has protested UAV attacks flown from inside Pakistan but may have used the U.S. presence as cover for Pakistani-flown attacks against terrorists. The United States has offered some concessions in UAV operations to mollify Pakistan but has not been willing to end the flights. According to press accounts, a U.S. Special Operations unit conducted a highly successful drone campaign against the leadership of the Islamic State (known as ISIS, ISIL, or the Daesh). The unit deploys Gray Eagle drones (formally, MQ-1C) that have a flight endurance of twenty-five hours and can carry four bombs or between four and eight missiles.

Another area of controversy in UAV operations has been civilian casualties. Although steps are taken to avoid or minimize these casualties, the fact that terrorists live within local populations makes these casualties inevitable given the number of UAV missions. There is a certain odd aspect to some of the protests over the use of drones and civilian casualties given the near certainty that using conventional munitions from airplanes or cruise missiles would put many more civilians at risk. This issue became more controversial in April 2015, when President Obama announced that a UAV strike on al Qaeda in Pakistan in January had also killed two Western hostages, Warren Weinstein and Giovanni Lo Porto; the United States subsequently gave Lo Porto’s family 1.185 million euros “in his memory.” On July 1, 2016, the Office of the DNI (ODNI) acknowledged that 64 to 116 noncombatants had been killed in 473 drone strikes in nonactive combat areas over the course of the Obama administration. Critics and private groups that track drone strikes questioned the official figures for noncombatant deaths. President Obama also signed Executive Order 13732 listing steps to minimize civilian casualties and calling for an annual report similar to the one released in July 2016. As noted in chapter 5, a drone strike during the U.S. evacuation from Kabul in August 2021 hit a car filled with civilians, including children, after it had been misidentified as a terrorist vehicle.

The greatest controversy about UAVs stemmed from the killing of Anwar al-Awlaki in a drone attack in Yemen in September 2011. Awlaki was born in the United States in 1971 and had dual U.S.-Yemeni citizenship. He apparently became radicalized in the 1990s—in part as a reaction to FBI surveillance that revealed Awlaki’s use of prostitutes—and, after leaving the United States in 2002, served as a recruiter for al Qaeda and encouraged terrorist attacks against Americans and American targets. Awlaki had been linked to Maj. Nidal Hasan, who attacked fellow soldiers at Fort Hood, Texas, in November 2009, and to Umar Abdulmutallab, who tried to set off explosives on an airliner in December 2009. The Fifth Amendment to the Constitution states that “[no] person [shall] . . . be deprived of life, liberty, or property without due process of law.” Thus, a question arose concerning the legality of the decision to kill Awlaki, which was a targeted strike. This debate became more pointed during the hearings over John Brennan’s nomination to be director of the CIA in early 2013.

Brennan, as President Obama’s counterterrorism adviser, had given a speech in April 2011 in which he acknowledged the targeted strike policy in Pakistan and alluded to it in Yemen and Somalia. In February 2013, the Justice Department released a white paper that offered a legal justification for using lethal force against a U.S. citizen operating abroad who was a senior operational leader of al Qaeda or an associated force. The main criteria are (1) that the individual poses an “imminent threat of violent attack” against the United States, (2) that “capture is infeasible,” and (3) that “the operation would be conducted in a manner consistent with applicable law of war principles.” According to press reports, President Obama reviewed the names of possible terrorist targets in a weekly “kill list,” as it was popularly called, but did not approve specific strikes. The Obama administration criteria included verifying the identity of a proposed target. There were two different types of drones strikes: (1) against specifically identified “high-value targets,” and (2) “signature strikes” based on observed activities—such as large groups of armed individuals or groups of vehicles—in areas where terrorists operate and that tend to indicate possible terrorist activity. According to press accounts, the CIA had been conducting drone strikes in Pakistan, but Obama shifted responsibility for the actual strikes from the CIA to the Defense Department.

In a May 2013 speech, President Obama admitted that four U.S. citizens had been killed in drone strikes, one alongside Awlaki and two others—including Awlaki’s son—in strikes on larger groups. In June 2014, a federal appeals court released portions of a 2010 Justice Department memo that gave the legal justification for the attack on Awlaki, laying out the arguments that were later embodied in the 2013 white paper. In November 2015, a federal appeals court ruled that internal documents concerning targeted operations against noncitizens could be kept classified.

Despite the Justice Department white paper, the policy of targeting certain U.S. citizens remains controversial. Former DCIA Michael Hayden pointed out an interesting irony when he observed that it takes a court order to tap the telephone of a U.S. citizen who is a terrorist, but not to kill him under certain conditions. Having previously fought successfully in court to maintain the secrecy of drone-related memos, in his May 2013 speech Obama announced the release of material that was briefed to Congress. He also announced new restrictions on the use of drones. As noted above, in July 2016, the Obama administration released figures for drone strikes and combatant and noncombatant deaths.

In November 2013, three Democratic members of the Senate Intelligence Committee supported the attack against Awlaki but called for a better understanding of the “evidentiary threshold” for conducting attacks on U.S. citizens. As of early 2014, the Senate and House Intelligence Committees were at odds on this issue, with the Senate favoring an additional review process for such attacks and the House opposing this. The Senate committee also supported an annual public report on the casualties resulting from drone strikes, which the House committee opposed.

Again according to press accounts, DCIA Mike Pompeo (2017–2018) argued successfully to President Trump that CIA authority to conduct drone strikes should be broadened to include Afghanistan, Yemen, and Somalia. Critics held that this decision would erode the greater transparency into drone operations that had been part of Obama’s order. The reported change in the CIA role regarding drones was consistent with other press reports that Pompeo had also been told to expand the role of CIA teams in Afghanistan, in conjunction with Afghan forces, to locate and kill Taliban militants.

In May 2021, Joe Biden’s administration made public a redacted version of the UAV strike rules in place during the Trump administration, in response to a lawsuit that overlapped both administrations. The Trump rules allowed field commanders some latitude within a set of “operating principles,” including the “near certainty” that there would not be civilian injuries or deaths. However, the Trump rules also allowed exceptions “where necessary.” Biden changed that policy in February 2021, as part of his review of counterterrorist policy. With the exception of strikes where U.S. military forces are in combat, meaning Afghanistan, the White House would now approve all other UAV strikes.

Several other issues also arose regarding the use of armed UAVs. One was the issue of congressional oversight, often expressed in the buzzword “transparency.” Although the intelligence committees are briefed on covert action findings, this may not be the case for UAV policy, which—as noted—is not necessarily a covert action as currently used. Members wanted more insight into the basis of the use of drones, leading to the release of the Justice Department memo. Some members also suggested establishing a secret court, similar to the Foreign Intelligence Surveillance Court (FISC), to review potential UAV attacks. Critics of this proposal pointed out that courts review legal cases but are not intended to make wartime decisions and that such a court could, in effect, become superior to decisions made by the president as commander-in-chief. Also, federal judges are not elected; the president is. Such a court would also not necessarily improve the due process rights of a proposed target as the target would have no representation in or chance to appeal the proceedings. In April 2016, a federal appeals court dismissed a suit seeking more access to how decisions are made to use drones.

Another issue was the question of which government agency or agencies should conduct drone strikes. They have been conducted by both the CIA and the Defense Department, although President Obama and DCIA Brennan both expressed the view that these flights should be shifted to Defense, in large part if not entirely. Such a shift would make the policy more transparent but would add other complications. The chain of command is shorter and more immediate in the CIA than it is in the military. Also, given that UAVs fly from bases in foreign countries, there is the issue of keeping the host government informed of operations, which is usual for U.S. military forces but not for CIA activities.

Despite this stated goal of shifting UAV-based attacks from the CIA to Defense, the transition apparently did not occur. In January 2014, the omnibus appropriations bill reportedly included language barring such a shift. Various senators complained that they were unaware of the language, which is contained in a classified annex. According to aides, Sen. Dianne Feinstein, D-CA, then chair of the Senate Intelligence Committee, and other legislators had concerns about how Defense would manage the UAV program and whether Defense would take steps to prevent collateral damage. Supporters of the current program also argue that the CIA has, by now, developed much greater expertise in these operations. Those who favor a shift to Defense argue that the emphasis on UAV operations is harming the core CIA roles of espionage and analysis.

ASSESSING COVERT ACTION

In addition to raising ethical and moral issues, the utility of covert action is difficult to assess. When examining a covert action, what constitutes success? Is it just achieving the aims of the operation? Should human costs, if any, be factored into the equation? Is the covert action still a success if its origin has been exposed?

Some people question the degree to which covert actions produce useful outcomes. For example, critics point to the 1953 coup against Iranian premier Mohammad Mossadegh and argue that it helped lead to the Khomeini regime in 1979. Proponents argue that an operation that put in place a regime friendly to the United States for twenty-six years, in a region as volatile as the Middle East, was successful. If no covert action is likely to create permanent positive change given the volatility of politics in all nations, is there some period of time that should be used to determine the relative success of a covert action?

As with all other policies, the record of covert action is mixed, and no hard-and-fast rules have been devised for assessing them. Assistance to anticommunist parties in Western Europe in the 1940s was successful; the Bay of Pigs was a fiasco. The view here is that the Mossadegh coup was a success, for the reasons noted earlier. But covert action is also subject to the law of unintended consequences. Abetting the fall of Allende helped lead to the regime of Gen. Augusto Pinochet. Average Chileans were probably better off than they would have been under an evolving Marxist regime, but many people suffered repression and terror. Aid to the mujahideen in Afghanistan was highly successful and played an important role in the collapse of the Soviet Union. At the same time, Afghanistan remained mired in a civil war ten years after the last Soviet troops withdrew and was eventually ruled by the Taliban, which hosted the al Qaeda terrorists.

This is sometimes seen as a “law of unintended consequences” effect. The fictional vaccination program created in Pakistan as a means to obtain DNA on the individuals in the bin Laden compound in Abbottabad is another example, given the subsequent Muslim militant attacks killing dozens of Pakistani public health workers, claiming they were spies (see chap. 5). Try as policy makers and covert actions planners might to foresee all of the possible outcomes and effects of an operation, this clearly is not possible. Indeed, it may become more problematic the further one gets from the covert action itself.

Covert action tends to be successful the more closely it is tied to specific policy goals and the more carefully defined the operation is.