CIA General Counsel Speech on Hypothetical Uses of Force

by Deborah Pearlstein

The speech delivered by CIA General Counsel Stephen Preston at Harvard yesterday is important and illuminating, and I agree with Ken the administration should be commended for it. But wow does it raise some troubling questions about how the CIA understands the legal authority for and constraints on its drone operations. There’s too much to unpack in it for one blog post, and I’d urge those who follow these interests to read it for themselves. Meantime, I’ll start with two issues: (1) the CIA’s understanding of its domestic authority to use force; (2) the CIA’s understanding of whether/how international law constrains its actions.

Domestic authority. Preston correctly explains that the CIA must have some source of authority under domestic U.S. law to carry out “hypothetical” activities involving the use of force abroad. In this inquiry, of course, international law is irrelevant. And I don’t read Preston to suggest that international law can give the U.S. government powers it does not otherwise possess under our own Constitution and laws. So what gives the CIA its authority to carry out drone strikes? Here’s Preston:

“First, we would confirm that the contemplated activity is authorized by the President in the exercise of his powers under Article II of the U.S. Constitution, for example, the President’s responsibility as Chief Executive and Commander-in-Chief to protect the country from an imminent threat of violent attack. This would not be just a one-time check for legal authority at the outset. Our hypothetical program would be engineered so as to ensure that, through careful review and senior-level decision-making, each individual action is linked to the imminent threat justification.

A specific congressional authorization might also provide an independent basis for the use of force under U.S. law.

In addition, we would make sure that the contemplated activity is authorized by the President in accordance with the covert action procedures of the National Security Act of 1947, such that Congress is properly notified by means of a Presidential Finding.”

Several points. Preston leads with – giving the impression that it does not only some but significant lifting in authorizing CIA actions – the President’s power under Article II of the Constitution. The non-reliance on Article II as an independent font of authority in U.S. counterterrorism operations – as opposed to statutory authorizations with specific limits – has been one of the central ways in which the Obama Administration has distinguished itself from the Bush Administration, which claimed sweeping authorities under Article II. So seeing it feature prominently here is striking. There are, important to emphasize, important differences between this invocation of Art. II power and Bush’s. Preston cites the President’s Art. II power to, as the framers put it, “repel sudden attacks.” There are few who would doubt the existence of such a power, and the Supreme Court has recognized it in various ways back to the Civil War era Prize Cases. This is in principle a narrower claim of authority under Article II than the (Bush) claim that Article II generally gives the President the power to detain people, interrogate them, and tap their phones as long as we’re in a state of armed conflict. How much narrower? Narrower at all? It depends a heck of a lot on what counts as “an imminent threat of violent attack.” How imminent does it have to be? Something more clear and specific than the general state of threat we face from, e.g., Al Qaeda? If a generalized threat from a group that’s attacked us at some point in the past is enough, then I confess I’m not sure how to distinguish this from the Bush-era understanding of Art. II.

Presumably Preston fronts the Art. II authority in part because CIA thinks that the statutory powers on the books don’t suffice to authorize all of the uses of force the CIA has carried out under its drone program. Indeed, the list of powers, Constitutional and statutory, is framed strangely. Preston says CIA first would make sure the action is authorized by the Constitution. And then notes, in quite different terms, “also” or “in addition” that statutory authority might exist. As if CIA doesn’t actually see the statutes as independent sources of authority. Do we take from this that CIA sees all of its use of force authority as coming from Article II, and the statutes on the books are just regulations, not themselves relevant sources of authority?

In particular, Preston doesn’t mention the 2001 AUMF expressly (which all 3 branches of government have interpreted as authorizing a global war against Al Qaeda). I can imagine 2 reasons why he might not want to get into the AUMF. One, the force authorized by that statute is limited to the groups who attacked us on 9/11, and some of the CIA’s targeting operations (it appears from press reports) have been aimed at individuals or groups who are only very arguably connected to those attacks. Is Al Shabab, for example, born as a domestic Somali insurgent group, really one of the organizations Congress meant to reach in its 2001 AUMF? Dicey to claim so. But we seem to be targeting some of their members anyway. Two, the administration has taken the position in court and elsewhere that international law, including the law of armed conflict, should inform the interpretation of (and, it would seem, constrain the use of) what force is authorized under the AUMF. But the CIA seems to have a different view of the applicability of LOAC, on which more below. So I can see why they wouldn’t want to rely on AUMF if they can help it. Which brings us to…

International law. Is there anything in international law – law of war or customary international law – that the CIA thinks it is bound to comply with as a matter of law (as opposed to, say, sensible policy or practice)? My read of this speech is that the answer is no. I would be very happy to be contradicted. The key sentences are described as a question of “compliance in execution with reference to international law principles.” And I’m not sure what the word “principles” is doing in there except to soften the notion that many of the relevant rules that might apply are simply law – indeed, when it comes to treaties the United States has signed and ratified, “supreme law of the land” under the Constitution. Here’s the relevant paragraph in its entirety:

“Here, the Agency would implement its authorities in a manner consistent with the four basic principles in the law of armed conflict governing the use of force: Necessity, Distinction, Proportionality, and Humanity. Great care would be taken in the planning and execution of actions to satisfy these four principles and, in the process, to minimize civilian casualties.”

If the past decade has taught us anything, I’d kinda think it’s this: Pursuing a policy “in a manner consistent with” the law is not the same as pursuing a policy that is bound by the law. So what’s going on? And now we enter the realm of pure speculation, but I guess that’s what blogs are for. So here’s what I imagine. I imagine that the CIA is targeting two kinds of people: (1) those it believes are participating in the armed conflict the United States has defined (i.e. a war against Al Qaeda and associated forces), and (2) those who are not plausibly understood as part of that armed conflict.

If the United States is targeting people in category (1), we are bound, as a matter of law, to comply with the law of armed conflict, which of course include, as a matter of law, the Geneva Conventions containing the rules Preston lists. We may well be complying with those rules – both the military and the CIA – that is, for example, not violating rules of proportionality in targeting. But even if we are complying with those rules – and boy do I wonder if and to what extent the CIA agents are trained in them – if agents of the CIA are pulling the trigger, I would think they may then be subject to criminal prosecution by domestic or foreign (or, if a tribunal with jurisdiction came to exist, international) courts for unlawful acts of violence they commit as unprivileged belligerents. CIA civilians are not members of our armed forces, and do not otherwise (as far as I know) meet the criteria under GCIII, Article 4 to lawfully participate in hostilities. So I can see why the CIA might be loath to acknowledge the applicability of these rules as law. But apply as law they do.

As a matter of international law – specifically, the UN Charter, to which we are a party, and which Preston cites – the United States may not lawfully target people in category (2) (i.e. those not part of our already quite broad armed conflict) unless it is exercising the “inherent right of individual or collective self-defence if an armed attack occurs …, until the Security Council has taken measures necessary to maintain international peace and security.” In the clearest summation of what the United States thinks that means that I’ve seen of late, Preston says that right includes, “for example, [where] the United States has already been attacked, and its adversary has repeatedly sought to attack since then and is actively plotting to attack again, then the United States is entitled as a matter of national self-defense to use force to disrupt and prevent future attacks.” Now there’s a ton to say solely on the question of whether this is a fair interpretation of the right of self-defense. But let’s assume for the moment it is. The rules that govern the exercise of that use of force in self-defense – i.e. how much force can you use, against whom, under what circumstances, etc. – are not only “principles.” They are customary international law, even by, I’d long thought, the estimation of our own government. So why not just say, at the very least, the CIA is bound by the customary international law of “Necessity, Distinction, Proportionality, and Humanity” governing the use of force? On this point, I have reached the limit of my imagination.

8 Responses

  1. Good post Deborah.  I have a lot of questions about this as well.  One relates to the domestic authorization bit, which you discuss.  I’m not as confident as you that Preston’s vague use of “specific congressional authorization” was meant to avoid the 2001 AUMF though I admit that it’s strange to not mention it if that’s what he meant.  I guess my question would be, assuming he’s not referring to the AUMF, what is he referring to?  It can’t be the 1947 NSA’s covert action authorization because Preston lays that out as requirement that exists “in addition” to authorization via Article II or “specific congressional authorization.”  Is there some other statute that could provide this kind of authorization?  If not, I think he probably is referring to the 2001 AUMF.  My speculation on why he leads with Article II authority though is basically the first point you make: some groups the CIA is targeting may fall outside of the 2001 AUMF.  However, some groups the CIA is targeting – such as AQAP – may fall into the administration’s claimed authority under the 2001 AUMF to target AQ (because they’ve become part of AQ) or “associated forces” of AQ (because they have assumed something akin to co-belligerent status), which is why the AUMF arguably might provide an “independent” basis for these groups but not other groups.  That has some pretty troubling implications.  But an even more troubling aspect of this speech is something that I think has been recurring problem in the administration’s legal views: implications for human rights constraints.  Specifically, when targeting what you call category (2) – those who are not plausibly part of an ongoing armed conflict – there seems to be no mention of the applicability of human rights law or principles as constraints on the use of force.  U.S. views on the extraterritorial application of HR law aside, that’s a very serious and profound mistake with regards to what law applies.  While the law of war conduct of hostilities rules may apply as baseline constraints to any use of force, it doesn’t follow that compliance with these rules is all that’s required (beyond meeting Article 51’s requirements).  Whatever one believes of the particulars of human rights constraints on the use of force in non-armed conflict scenarios, they are certainly more stringent than the use of force constraints applicable in armed conflict.  It might be that the USG’s view of what is required under an constitutional Article II analysis plus Article 51 analysis plus law of war use of force constraints will always equal the constraints imposed by human rights on the use of force, but I doubt it.  I doubt it in part because neither the constitutional Article II analysis nor the Article 51 analysis are designed to get at human rights obligations (interestingly, the constitutional due process analysis Holder laid out vis-a-vis citizens gets closer), but also because of the very serious open questions regarding how the USG interprets the imminence requirement.

  2. Response…
    But, Raha Wala, it has been mentioned here before that, yes, human rights law applies universally (e.g. U.N. 55(c), 56; ICCPR, prmbl.; CAT) and in all social contexts (e.g., also in time of war), but who is protected and, then, with what protections?  With respect to “who,” outside the territory of the U.S., for example, those who are in the actual “power or effective control” of the U.S. (not those being targeted by high flying drones).  Perhaps for this reason, there was inadequate attention to human rights law, but the point could have been made.  With respect to h.r. protections, the ICCPR prohibits “arbitrary” death.  Lawful targetings during war or self-defense under UN art. 51 would not be “arbitrary.”  This point should also have been made.

  3. Good points Jordan.  I suppose I should have lumped in the effective control issue with the extraterritorial application issue (clearly they are interrelated), but I’m not sure I would concede that drone attacks are outside of the jurisdiction of HR law.  That’s in part because I might have a more elastic view than you of what effective control means (physical control could be sufficient for effective control, but might not be necessary).  I also wouldn’t say that compliance with art. 51 means that the deprivation of life isn’t “arbitrary” per se since art. 51 governs the interstate use of force but doesn’t go all the way to answering whether any particular use of force against an individual or group is lawful under international law and therefore not arbitrary.  Even with compliance with art. 51, a particular use of force could be unlawful if it, for example, violated the principle of humanity.  One of the my quibbles is that, as far as I can tell, the government’s view is that once you pass through the art. 51 gate, at most the law of war conduct of hostilities rules apply as a further limit on the use of force (though possibly the gov’t doesn’t even think they apply as a matter of law, as Deborah points out), even in a non-armed conflict scenarios.  I want to say that at minimum those rules apply, but there are further constraints applicable by reference to HR law in non-armed conflict scenarios.

  4. Response…
    Yes, we seem to disagree with respect to the “effective control” test.  I recall the general debate about bin Laden’s death for example.  I have concluded that he was not in the “effective control” of the U.S. Navy Seals team.  Under the laws of war, he would have had to have “surrendered” — HC IV, Annex, art. 23(c).
    With respect to self-defense targetings, it may very well be that CIA GC Preston has adopted what I term use of the movie camera instead of a snapshot (see the cite above) or a process of armed attacks focus — and, clearly, al Qaeda has been directly involved in continual armed attacks, a process of armed attacks, on U.S. military and other personel in Afghanistan and elsewhere.  Therefore, the U.S. has a continual right of self-defense to target those who are DPAA (directly participating in armed attacks). 
    Also, it is logical and policy-serving to use, as Preston states, “the four basic principles in the law of armed conflict governing the use of force: Necessity, Distinction, Proportionality, and Humanity” when engaging in lawful self-defense targetings under UN 51, since, my gosh, I wrote similarly, self-defense targetings are legally conditioned by general principles of reasonable necessity and proportionality.  see cite above and
    EEgads — are CIA lawyers, including the GC, reading my stuff?
    With respect to domestic authority, since treaties are part of the “Laws” of the United States and are binding on the President and all members of the Executive branch (along with customary international law — see ) presidential power is also enhanced in the sense that the President has the constitutionally-based authority (and a responsibility) under Art. II of the Constitution, as the Executive and with respect to the faithful execution of the Laws clause, to execute the competence of the United States under Article 51 of the U.N. Charter to engage in otherwise permissible self-defense measures against those who are engaged in ongoing armed attacks on the United States and its nationals abroad.  Wasn’t that what GC Preston had in mind?

  5. Great post and great comments.  Raha, I think there is no doubt that your view of effective control is more elastic that that of the US government (USG).

    On the substantive point of the applicability of the principles of international humanitarian law to CIA lethal force operations, I will quote — well — myself [with minor clarifications/additions], from an earlier comment I made to a post by Ken Anderson on this topic:

    “…My read of official statements and litigation positions [and the CIA General Counsel’s speech], and I think Kevin’s  [and Deborah’s] too, is that the USG assertion is that [CIA] attacks are part of the AUMF non-international armed conflict (NIAC) or are independent lawful acts of national self-defense unrelated to that conflict.  This is an important distinction….

    To successfully support such attacks as an independent act of self-defense, the USG must necessarily believe that IHL applies (perhaps only by analogy) to acts of self-defense against non-state actors even if hostilities with them do not meet the oft-asserted customary international law (CIL) duration, intensity and organization thresholds to establish a NIAC.  I think this is what Marty [Lederman] earlier stated as his understanding of the USG view, and Michael Lewis argues to be state practice as well.
    Kevin [Jon Heller]’s (and Mary Ellen [O’Connell]’s, [and perhaps Raha’s] and others’) disagreement with this is that it conflates the jus ad bellum and jus in bello for NIAC.  They think even if there is a jus ad bellum right of self-defense against a non-states actor, a state is limited to IHRL (law-enforcement methods) unless hostilities in a given geographic location meet the intensity, duration and organization requirement[s].  (So I guess this means that if a non-state actor projects its attacks far enough away, geographically and temporally disperses them, and keeps itself loosely- or un-organized, it can effectively prevent the application of IHL no matter how much damage it inflicts.)

    In recent commentary (e.g. our PENNumbra debate and here), it seems Kevin would allow attacks beyond active conflict zones, but only if the targets’ activities give them targetable status in IHL and are directly related to the hostilities occurring in an active conflict zone.  Mary Ellen would not permit that position, but recently stated that she would permit an intrusion upon a foreign state’s territorial integrity to engage in extraterritorial law enforcement, the intrusion necessary to do so being justified as a counter-measure….  

    If I accurately state the USG view of the law governing self-defense not involving hostilities reaching the NIAC threshold, then it may be based in a more holistic view of armed conflict and the applicability of IHL.  If so, the USG would be required to state that IHL…[principles apply] to all armed conflict, and that governments almost necessarily determine when such conflicts exist by resort to armed attacks [against non-state actors].  Alternatively, the USG could state that neither IHL nor IHRL apply to these acts of extraterritorial self-defense, but that when undertaken in response to actual or imminent attacks of sufficient intensity, it is good practice to observe IHL requirements of necessity, distinction and proportionality [and humanity].  On the whole, I think this [latter view] may be the best reading of Koh’s statements and the USG’s actual litigation positions [as well as the CIA General Counsel’s recent remarks].”

  6. John, thanks for that helpful post.  Just a point of clarification: while my position is that the intensity, duration, and organization requirements must be met for IHL to apply, I don’t necessarily subscribe to geographic limit view.  (Though query whether certain geographic elements might indirectly factor into the aforementioned three requirements.)

    You pose the interesting hypothetical of a non-state armed group that (intentionally) stays un-organized or loosely-organized, temporally disperses its attacks, and launches them from far enough away – all to avoid the application of IHL requirements.  I doubt this hypothetical actually matches up with the social science of how non-state armed groups behave so its unclear what value it has in informing ongoing legal and policy debates.  Anyhow, I don’t think the hypothetical is as fatal to what I’ll call the human rights position as you imply.

    As an initial matter, I disclaimed the geographic limit perspective so non-state armed groups can’t just escape off to any part of the world to avoid the application of IHL.  The rest of the requirements – a sufficiently organized armed group and hostilities of sufficient intensity and duration – make good sense as line drawing to ensure that the laws of war, which are supposed to apply in exceptional circumstances, don’t overtake and overwhelm peacetime rules.  Nor does the application of human rights law mean that a state is powerless to act in the face of a menacing threat.  Even in peacetime, where there is an imminent threat to civilians and no another way to deal with it, the use of deadly force is often permitted.

    Another way to think about this is what the world could like without the threshold requirements for the application of IHL.  After all, a lot of people cause a lot of “damage” – as you put it John.  Gangs kill hundreds if not thousands a year.  As do drug cartels.  And there are many more terrorist groups than al Qaeda and associated forces; the majority of them are not motivated by violent Islamist ideology and routinely carry out plots domestically.  If governments simply determine when armed conflicts exist by resorting to armed attack to switch on IHL rules, we have the flip-side of your hypothetical; governments now have the incentive to use force, create armed conflicts, and claim the more permissive detention and targeting rules of IHL in situations that simply do not pose the exceptional and grave threats associated with armed conflict.

    So assuming that there is agreement IHL should only apply as an exceptional matter in situations of armed conflict, I think the burden is on those who don’t like the CIL threshold requirements for armed conflict to identify some other line-drawing factors or requirements that make sense.  I don’t think it’s good enough to just say that states decide by political means or resort to force – that may make sense as a matter of domestic law but I don’t think it fits in with any coherent theory of international law.

  7. I don’t disagree with you at all, Raha.  I only meant to survey the landscape of opinion and postulate the US view.  I simply think that the USG view being articulated over and over again is that IHRL does not apply (extraterritorially), and that compliance with IHL principles is sound policy, but not absolutely legally required.  I don’t agree with this view as a legal matter.

    My point regarding loose organization and geographical dispersion was to observe that is a possible tactic that may be employed by sophisticated groups, not that it necessarily is.  I continue to believe that in an an age of low-cost, high-tech gadgetry which can be used to inflict great harm the view of states regarding armed conflict threshold might shift and possibly alter CIL.  Some seem to argue that it already has.  I don’t think a different CIL standard has formed on this point, but it seems to be moving that way. Again, if the U.S. view is that resort to armed force is enough to trigger IHL, it should clearly say so. I wouldn’t agree with that position as a legal matter, either.

  8. Thanks John.  I think you’re probably right about the U.S. view, though I do at times find it difficult to discern 1) what the sources of legal authority and constraints are for particular USG actions and 2) with respect to the restrictions, whether they are being followed as a matter of law or policy.
    As you note, the US view on extraterritorial application of IHRL has historically been more clear than its view on IHL and other international law regulations.

Trackbacks and Pingbacks

  1. There are no trackbacks or pingbacks associated with this post at this time.