Archive for
December, 2017

One Step Forward for International Criminal Law; One Step Backwards for Jurisdiction

by Jennifer Trahan

[Jennifer Trahan is Associate Professor, The Center for Global Affairs, NYU-SPS and Chair of the International Criminal Court Committee of the American Branch of the International Law Association]

On Thursday, December 14, 2017, the ICC’s Assembly of States Parties (ASP) took the historic and significant decision, by consensus, to activate, effective July 17, 2018, the ICC’s jurisdiction over its 4th crime, the crime of aggression. (The Kampala crime of aggression amendment had been “adopted” in 2010 at the Kampala Review Conference, but there was a delay mechanism such that jurisdiction did not yet “activate”, but first required 30 States Parties to ratify the amendment (35 now have), and one more decision by the ASP to activate.)

The decision made by the ASP was a step forward for international criminal law, a step forward for completing the Rome Statute as envisioned in 1998 (which already included jurisdiction over 4 crimes), a step forward for carrying on the legacy of the International Military Tribunal at Nuremberg, and a step forward in trying to create more deterrence behind UN Charter article 2(4). But, it was a step backwards in how to read the Court’s jurisdiction over the crime.

While many had hoped that at the ASP, it could be agreed to simply “activate” jurisdiction by consensus (for instance, simply reflected in a sentence in a resolution), already over the past year it appeared that would not be the case. As many readers will know, there have been two different readings of what was accomplished in Kampala.

The differences in reading pertained to which States Parties would be covered by the ICC’s crime of aggression jurisdiction after activation in the situation of State Party referral or proprio motu initiation of investigation (Rome Statute article 15 bis). (Activation also triggers the possibility of UN Security Council referrals covering the crime of aggression (article 15 ter); non-States Parties were completely exempted from the crime’s jurisdictional reach already during the Kampala negotiations (art. 15 bis, para. 5).)

One reading (let us call it the Liechtenstein/Swiss/majority reading) was that after the activation decision, for purposes of State Party referrals and proprio motu initiation, ALL States Parties could be subject to crime of aggression jurisdiction, absent their lodging an “opt out” declaration, but only also if either the aggressor or victim State Party had also actively ratified the crime of aggression amendment. The other reading (let us call it the UK/French reading), was that no State Party could be covered by the crime of aggression after activation unless it had also actively ratified the amendment. (This reading results in an extremely restrictive jurisdictional regime, because, frankly, ratifying States Parties such as Liechtenstein and Botswana are not invading each other.)

After a year of a “facilitation” process, led by Austria, to try to resolve this issue, negotiations opened during the ASP. What I am calling the Liechtenstein/Swiss/majority group proposed various draft texts that could have helped bridge the gap between the two readings, with Brazil and Austria also proposing helpful suggestions. Yet, the UK/France (at times joined by Norway, Japan, Colombia, Australia, Canada and Denmark) insisted on their view simply prevailing, and, in the end, the UK and France never moved from that position. (This narrative reflects my understanding of negotiations gleaned from discussions with representatives of States Parties, as, unfortunately, members of civil society were excluded from the “closed door” negotiations.)

The desire to achieve consensus activation (meaning any State Party could block consensus) provided any single State Party (or two States Parties, as was the case here) with enormous leverage. A vote would require 2/3rd of States Parties voting for it, and a few delegations did not stay to the close of the ASP (so the full 123 States Parties were not present towards the end of the conference). Additionally, a vote suggests a divided commitment that States Parties did not appear to want, and how the vote would turn out seemed uncertain as well.

In the end, States Parties had (for many of them) the very difficult decision to make—whether to activate the crime in an historic and important decision, if it meant accepting the extremely restrictive reading of jurisdiction given by the French/UK group. (This is quite ironic because it means that the 4 states that had conducted the Nuremberg prosecutions are either now caved out of crime of aggression jurisdiction (the US and Russia as non-States Parties) or can easily do so by not ratifying the amendment (the UK and France).) On the other hand, a decision not to accept the UK/French reading meant that the negotiations would conclude with no agreement, and no clear commitment when, where or whether to resume negotiations, and no certainty that any resumed negotiation would conclude any differently in the future.

This author believes States Parties made the right decision. It was not what many of them had wanted and thought they had negotiated in Kampala. Yet, international law often moves forward in imperfect ways (the war crimes amendment also adopted at this ASP dropped a key war crime along the way). And, really, in the end of the day, all States Parties agreed that the crime of aggression is a consensual regime—and it was only how to achieve that (basically an “opt in” or “opt out” approach).

It was a large concession, which now means, at present when the crime of aggression activates on July 17, 2018 (the activation date selected in the activating resolution, ICC-ASP/16/L.10*), it will have extremely limited jurisdictional reach. The good news is the ICC will hardly be overwhelmed with cases (for those who worried about this)—it could even take years before there is a case of aggression within its jurisdiction. The bad news of course if that if one hoped the activation of the crime could have some deterrent impact in trying to prevent aggressive uses of force, including war, that deterrent impact is now lessened. (Deterrent impact is more likely now to be created through the possibility of U.N. Security Council referral—which could cover States Parties (whether or not they ratify) and non-States Parties). In terms of increasing the jurisdictional reach for purposes of non-Security Council referrals, it is now up to the ICC, civil society and States Parties to press for additional crime of aggression ratifications.

The Draft Resolution’s Curious Paragraph 3

by Kevin Jon Heller

A friend who is even more jaded than I called my attention to the following curious paragraph in the Draft Resolution the ASP has just adopted by consensus:

3.    Reaffirms paragraph 1 of article 40 and paragraph 1 of article 119 of the Rome Statute in relation to the judicial independence of the judges of the Court.

This paragraph is new — it was not included in the earlier Draft Resolution I blogged about. For those of you who are not total Rome Statute nerds, here is the text of the two referenced articles:

Art. 40(1): “The judges shall be independent in the performance of their functions.”

Art. 119: “Any dispute concerning the judicial functions of the Court shall be settled by the decision of the Court.”

I think my friend is right: Paragraph 3 likely represents the last gasp of the opt-out camp — a shameless plea to the judges to ignore the text and drafting history of the Draft Resolution and require states that have not ratified the aggression amendments to opt-out. Fortunately, as jaded as I am about the ICC’s judges, I think the likelihood of the plea ever succeeding is essentially zero. The text and drafting history are too clear. Moreover, a decision to adopt the opt-out position despite the text and drafting history of the Draft Resolution would be catastrophic for the Court. It would be bad enough if the OTP brought aggression charges against a state party that had not ratified the amendments. It could be the end of the Court — and I am not being Chicken Little here — if the judges permitted such charges to proceed. Such a decision could easily lead to the UK, France, Japan, and others to withdraw from the Court. And they would be justified in doing so.

The judges’ relentless judicial activism has damaged the Court enough. If the Court is to have any future — one in which states cooperate with it and use their muscle to ensure that it succeeds — states have to be confident that the judges will respect their will, even when that will is less than ideal.

Paragraph 3 should never have been included in the Draft Resolution.

ASP Adopts the Aggression Amendments by Consensus

by Kevin Jon Heller

It went down to the wire, but it’s over. States reached consensus on adopting the aggression amendments — after those in the opt-out camp gave in to the opt-in camp. The adopted Draft Resolution provides the following:

Confirms that… in the case of a State referral or proprio motu investigation the Court shall not exercise its jurisdiction regarding a crime of aggression when committed by a national or on the territory of a State Party that has not ratified or accepted these amendments.

This language is unequivocal, going well beyond the Draft Resolution I referenced in my previous post. Under the adopted Resolution, state parties do not have to do any in order to remain outside the Court’s aggression jurisdiction. Unless a state party ratifies or accepts the aggression amendments, it will be in the same position as a non-state party.

Having received a few rather nasty emails regarding my defense of the opt-in position, I want to make my substantive views clear. Although I completely agree with the opt-in states that, as a matter of treaty law, they could not be subjected to the Court’s jurisdiction over aggression in any way unless they ratified the aggression amendments, that is not my preferred jurisdictional regime. On the contrary, I believe that aggression should be governed by the same regime — automatic jurisdiction — that applies to the other core crimes. In particular, I strongly dislike the decision to exempt non-states parties from the Court’s jurisdiction even when one of their nationals commits the crime of aggression on the territory of a state party. I see no reason why state parties should not be protected against aggression by non-party states in the same way they are protected against war crimes, crimes against humanity, and genocide.

My reservations aside, this is clearly an historic day. Kudos to all the states, NGOs, and individuals — I am so glad the inestimable Ben Ferencz lived to see this — who made the activation of aggression possible.

A Different View of the Aggression Activation Negotiations – A Perspective from the Ground

by Gregory Gordon

[Gregory Gordon is Associate Professor of Law at the Chinese University of Hong Kong]

I’m here on the ground in New York and I want to provide an additional perspective on Kevin’s post. First, the document he has posted is strictly a transitory draft meant only to facilitate discussion. In no way does it necessarily represent any hardened position of the delegates. From what I understand, there is no consensus on it and the parties have moved on from it.

Second, even if it did represent the final word, his understanding about OP(1)(a) and (b) is not necessarily accurate. ICC-ASP/16/L.9/Rev.1 OP (1)(a) only acknowledges the position of those States Parties (i.e., the UK/France camp), which is already reflected in the Report on the facilitation, statements made here upon adoption of any resolution, or by letter to the ASP by 31 December 2018. Subsection (b) applies only to the States Parties referred to in (a). Kevin provides his interpretation that “they” extends to all other States Parties. But the understanding of folks here on the ground is that (b) is limited to those States Parties in (a).

Kevin is entitled to his interpretation — but it does not seem to accord with the tenor of negotiations. Moreover – and this is very important – and in all due respect to Kevin – at this point, it is not helpful to speculate on either the process or substance of negotiations that remain active and have a way to go prior to a final decision (either way) by States Parties. I appreciate that he is trying to stimulate interested discussion – that’s his right – but given the delicate nature of negotiations, public comment could have a potentially misleading and unintended detrimental effect. ICC-ASP/16/L.9/Rev.1, which Kevin has posted in its entirety, was not intended to be publicly released. There are good reasons for that. Many of us here are hoping that this delicate negotiation process will lead to an outcome that will contribute toward the progressive development of international criminal law. Let us hope those delicate negotiations – still very much in train — come to fruition.

The Opt-Out Camp Possibly Folds — Clearing Way for Aggression?

by Kevin Jon Heller

A new document is being circulated at the Assembly of States Parties entitled “Draft Resolution: Activation of the jurisdiction of the Court over the crime of aggression.” Operative Provision 1(b) seems to indicate that the opt-out camp, led by Liechtenstein, has conceded the jurisdictional point to the opt-in camp, led by Japan, Canada, and the United Kingdom. Here is the text of OP1(b):

(b)    The Assembly unanimously confirms that, in accordance with the Rome Statute, in case of a State referral or proprio motu investigation the Court shall not exercise its jurisdiction in respect of the crime of aggression when committed by nationals or on the territory of the States Parties referred to in subparagraph (a), unless they ratify or accept the amendments regarding the crime of aggression.

The provision makes clear that the Court will have no jurisdiction over any act of aggression involving a state party that does not ratify or accept the aggression amendments — thus placing states parties in the same position as non-state parties.

There is, however, one twist. To take advantage of OP1(b), a state will have to make its agreement with the opt-in camp known by no later than 31 December 2018, when the Court’s jurisdiction will begin. That’s the result of reading OP1(b) in conjunction with OP1(a). Here is the text of the latter provision:

(a)      The Assembly acknowledges the positions expressed by States Parties, individually or collectively, as reflected in the Report on the facilitation or upon adoption of this resolution to be reflected in the Official Records of this session of the Assembly or communicated in writing to the President of the Assembly by 31 December 2018 that, for whatever reason, including based on paragraph 5 of article 121 of the Rome Statute, they do not accept the Court’s exercise of jurisdiction over the crime of aggression unless they ratify or accept the amendments regarding the crime of aggression,

I can’t see why the Draft Resolution would not satisfy the opt-in states. So if the opt-out camp supports the resolution, it should ensure that the aggression amendments are adopted by consensus later today.

NOTE: As I read the Draft Resolution, states that join the Court after 31 December 2018 would have to opt-out of aggression jurisdiction, because OP1 would not apply to them. That’s an interesting twist/compromise!

The Puzzling US Submission to the Assembly of States Parties

by Kevin Jon Heller

The US submission to the ASP has finally appeared. It is not very long — about 1.5 pages — but manages to pack in a good number of false claims and bizarre interpretations of the Rome Statute.

In terms of falsity, the US repeats its longstanding claim that the Court has no jurisdiction over the nationals of non-state parties, even when those nationals are responsible for an international crime committed on the territory of a state party (p. 1):

As an initial matter, and as we have consistently emphasized, the United States is not a party to the Rome Statute and has not consented to any assertion of ICC jurisdiction, nor has the Security Council taken action under Chapter VII of the UN Charter to establish jurisdiction over U.S. personnel. It is a fundamental principle of international law that a treaty is binding only on its parties and that it does not create obligations for non-parties without their consent. The Rome Statute cannot be interpreted as disposing of rights of the United States as a non-Party without U.S. consent.

This is wrong, for reasons Dapo Akande has patiently explained. It’s also completely hypocritical, because the US had no objection to the Special Court for Sierra Leone (SCSL) prosecuting Charles Taylor, a Liberian national, even though the SCSL was created by an international agreement — between the UN and Sierra Leone — to which Liberia was not a party. Indeed, the current US submission emphasises that it was “one of the most vocal supporters for the creation of tribunals to try those most responsible for atrocities committed in Rwanda and Sierra Leone.”

The most bizarre argument in the submission has to do with the principle of complementarity (p. 1):

Additionally, we are concerned about any ICC determination — as required by the Rome Statute’s core principle of complementarity — on, for example, the genuineness of U.S. legal proceedings without United States consent. The principle of complementarity fundamentally limits the ICC’s exercise of jurisdiction to those cases in which a State is genuinely unwilling or unable to comply with its duties, such as those under the Geneva Conventions, to investigate and prosecute war crimes, genocide, and crimes against humanity. Just as we have not consented to jurisdiction over our personnel, we have not consented to the ICC’s evaluation of our own accountability efforts.

This is literally nonsense. The ICC would only formally assess complementarity in the context of a specific prosecution of an American national — and would only do so (practically, if not because of a legal limitation) if the US decided to challenge the admissibility of a case. So the US would have to “consent” to the Court examining the genuineness of American proceedings if it wanted to head off a prosecution. Beyond that, consent has nothing to do with complementarity.

I will avoid making snarky comments about the US’s claim (p. 2) that it “has undertaken numerous, vigorous efforts to determine whether its personnel have violated the law and, where there have been violations, has taken appropriate actions to hold its personnel accountable.” But I can’t let the following claim (p. 2) go unremarked:

Indeed, we note the irony that in seeking permission to investigate the actions of U.S. personnel, the Prosecutor appears to have relied heavily upon information from investigations that the United States Government itself decided to make public. We question whether pursuing this investigation will make other countries less willing or able to engage in similar examinations of their own actions and to be transparent about the results.

This is, well… ironic. The OTP’s request to open an investigation into Afghanistan notes multiple times (see para. 27 for an example) that the US refused to cooperate with the preliminary examination. And the request relies very heavily on the executive summary of the Senate Intelligence Committee’s “Study of the Central Intelligence Agency’s Detention and Interrogation Program” — which the White House and CIA tried desperately to keep from ever seeing the light of day.

All that said, I am delighted by the following statement in the US submission (p. 1; emphasis mine):

The principle of complementarity fundamentally limits the ICC’s exercise of jurisdiction to those cases in which a State is genuinely unwilling or unable to comply with its duties, such as those under the Geneva Conventions, to investigate and prosecute war crimes, genocide, and crimes against humanity.

The US has now formally acknowledged that it has a duty under international law “to investigate and prosecute war crimes, genocide, and crimes against humanity”! That is a bold and progressive claim, especially with regard to crimes against humanity, for which there is no treaty that demands either investigation or prosecution. I imagine that position will come as something of a surprise to the parts of the US government that were not involved in drafting the submission…

The US’s ASP submission: wrong, bizarre, but surprisingly — and probably inadvertently — progressive.

Against (False) Consensus — the ASP and the Aggression Amendments

by Kevin Jon Heller

Although many important issues will be discussed this week at the Assembly of States Parties (ASP), none will be quite so momentous as the decision to activate the ICC’s jurisdiction over the crime of aggression. Whatever one thinks of the merits of the definition of aggression, there is no question that the activation of jurisdiction will represent the culmination of seventy years of efforts to deem aggression an international crime.

When the ASP finally makes a decision concerning the aggression amendments, the Rome Statute will encourage it to do so by consensus. The relevant provision is Art. 112(7):

Each State Party shall have one vote. Every effort shall be made to reach decisions by consensus in the Assembly and in the Bureau. If consensus cannot be reached, except as otherwise provided in the Statute:

(a)     Decisions on matters of substance must be approved by a two-thirds majority of those present and voting provided that an absolute majority of States Parties constitutes the quorum for voting;

(b)     Decisions on matters of procedure shall be taken by a simple majority of States Parties present and voting.

With regard to amendments, the Rome Statute does “otherwise provide.” According to Art. 121(3), “[t]he adoption of an amendment at a meeting of the Assembly of States Parties or at a Review Conference on which consensus cannot be reached shall require a two-thirds majority of States Parties.” Similarly, Art. 123(3) says that “[t]he provisions of article 121, paragraphs 3 to 7, shall apply to the adoption and entry into force of any amendment to the Statute considered at a Review Conference.”

The aggression amendments were adopted by consensus at the 2010 Review Conference held in Kampala. That was a mistake, because there was no genuine consensus at Kampala concerning one particularly critical issue: whether the Court will have jurisdiction over an act of aggression committed on the territory of a state party that has ratified the aggression amendments by a national of a state party that has not ratified them. States have taken diametrically opposed positions on that issue. Most — led by Liechtenstein — believe that the Court will have jurisdiction in that situation unless the non-ratifying state formally opts-out of the crime of aggression. But some — led by Japan, Canada, and the United Kingdom — insist that the Court will have no jurisdiction because non-ratification is enough.

I don’t want to re-litigate the merits of the debate. Regular readers know I agree with the Japan/Canada/UK group. My point is more modest: to call attention to the danger that false consensus poses to the legitimacy of the crime of aggression. By adopting the aggression amendments by consensus, instead of through a formal vote, the ASP made possible the kind of bitter disagreement that has characterized the state-party opt-out/opt-in debate. Just consider the following paragraphs from the ASP’s most recent report concerning the facilitation of aggression’s activation (emphasis mine):

19. Some delegations suggested that a declaration lodged with the Registrar, indicating that a State Party did not accept the jurisdiction of the Court over the crime of aggression, would bring the desired clarity. These delegations stressed that the negotiating history of the amendments during the Special Working Group on the Crime of Aggression as well as during the Review Conference offered clear evidence of the correct legal interpretation of the agreement reached in Kampala… By enabling States Parties to declare that they do not accept the jurisdiction of the Court in respect of crimes of aggression committed by their nationals (“opt-out”), the compromise agreed in Kampala had chosen the middle-ground between two opposing positions. It was explained that the Review Conference had considered the opposing position papers concerning the second sentence of article 121, paragraph 5, (the so-called “positive” and “negative” understandings), which both were ultimately deleted in favor of including the opt-out regime. As a result, in their view the compromise reached at the Review Conference was clear, and jurisdiction of the Court could extend to nationals of those States Parties which had not ratified the amendments, unless they opted out.

20. Other delegations took the view that too much emphasis was being placed on the negotiating history of the amendments and on concessions or compromises, rather than on legal principles and the plain meaning of the texts. These delegations explained that they had left the Review Conference with a different understanding, namely that the amendments would not apply to those States Parties which would not ratify them. In their view, open legal questions as to the implications of activation remained. These had not been solved during the seven years since the Review Conference, in spite of ratifications by 34 States Parties. It was pointed out that negotiations at the Review Conference could not have the effect of changing treaty rights and obligations. Accordingly, it was important to focus on the ordinary language of the text as there were differing understandings of the negotiating history, since it was also possible to explain the opt-out as something that was available to ratifying States Parties. These delegations questioned how they could be required to take action to opt out of the Court’s jurisdiction if they had not chosen to opt in by ratifying the amendments in the first place.

This is false consensus, not genuine consensus. State have been debating the opt-out/opt-in point for seven years as a result of the “consensus” at Kampala — debate that has overshadowed the importance of the aggression amendments themselves and has cost the ASP significant time that could have been better spent debating other issues.

Both camps deserve blame for this situation. The “opt-out” camp deserves blame for continuing to insist that genuine consensus existed at Kampala even though it’s quite clear it did not. And the “opt-in” camp deserves blame for permitting the aggression amendments to be adopted by consensus even though they knew disagreement existed about such a fundamental issue.

It will be interesting to see what happens this week at the ASP. There are rumblings — according to my well-connected friend Don Ferencz — that the opt-in group might put up a fight:

The member states of the Court will meet at UN headquarters in New York from December 4th to 14th to address making good on the pledge which they made in Kampala. Although 34 nations have already ratified their acceptance of the Court’s aggression jurisdiction – including over half of the members of NATO – Britain and France have not. Instead, they have joined with a handful of states, including Japan, Canada, Norway, and Colombia, in tacitly threatening to defeat activation of the Court’s aggression jurisdiction by their insistence that the Court must first clarify that the aggression amendments will not apply to leaders of any state that does not independently ratify them…

It is significant that the upcoming decision on aggression is expected to be undertaken pursuant to a consensus resolution. This means that the activation resolution must either be adopted by unanimous approval or not adopted at all. In such circumstances, each member state of the Court has the power to thwart the will of even an overwhelming majority simply by not consenting to the adoption resolution, regardless of the express terms of what was unanimously agreed to in Kampala. The non-ratifying countries which are demanding clarity that their leaders will remain beyond the Court’s reach on the crime of aggression, therefore, each have a potentially game-ending card to play in opposition of the final approval. The question is, with the whole world watching, do they dare play it?

My personal hope is that the opt-out camp will give in and accept the opt-in camp’s position that the Court will have no jurisdiction over a state party that does not ratify the aggression amendments — thereby creating genuine, if forced, consensus. Not only do I think that opt-in is the correct legal position, I am very skeptical that the opt-out camp would have enough votes to adopt the aggression amendments over the objections of the opt-in camp. According to paragraph 28 of the ASP’s facilitation report, they would need 82 in the absence of consensus. And that seems unlikely, given what I’ve heard about the number of states that are either opt-in or plan to abstain on any vote to activate the Court’s jurisdiction over aggression.

UPDATE: I have updated the post to mention Arts. 121(3) and 123(3) of the Rome Statute, as well as my understanding of the state of play concerning the vote to activate jurisdiction.

Reminder: Seventh Annual Junior Faculty Forum for International Law

by Kevin Jon Heller

Reminder: Deadline for Applications for Seventh Annual Junior Faculty Forum for International Law—May 28, 29 and 30, 2018

This is a brief reminder that applications for the Seventh Annual Junior Faculty Forum for International Law are due on Dec. 15, 2017. The Forum will be convened by Anne Orford (Law – Melbourne), Dino Kritsiotis (Law – Nottingham) and J.H.H. Weiler (Law – NYU) and will be held at the University of Melbourne in May next year. Full details of the application process are available on the Forum website (http://annualjuniorfacultyforumil.org/). Applications are welcome!

Cyber Operations and GCII Article 18’s “End of Engagements” Clause

by Jeffrey Biller

[Jeffrey Biller, Lt Col, USAF, is the Associate Director for the Law of Air, Space and Cyber Operations at the Stockton Center for the Study of International Law, US Naval War College.]

On 27 May 1941, the British battleships King George V and Rodney engaged the German battleship Bismarck, which had been previously disabled by a torpedo attack from aircraft belonging to the British carrier Ark Royal. After almost two hours of fighting, the Bismarck and her 2200 man crew were sunk. As the Bismarck’s escort ship, the Prinz Eugen, had previously detached, the shipwrecked crew was entirely dependent on the Royal Navy for rescue. The British ships Dorsetshire and Maori, acting in accordance with Article 16 of the Convention (X) for the Adaptation to Maritime Warfare of the Principles of the Geneva Convention, began rescue of the German crew. However, after 110 sailors were rescued a U-boat alarm was sounded, forcing the Royal Navy to break off the rescue. All but five of the remaining German crew were lost at sea.

The obligation under which the British acted to rescue the crew of the Bismarck was expanded in Article 18 the Second Geneva Convention (GCII). In an earlier post, Lt Cdr Peter Barker, RN, analyzed the extent of the obligation placed upon warship commanders to search for and collect the shipwrecked, sick, and wounded following a naval engagement. He correctly identified that the law, contained primarily in Article 18 of GCII, is ambiguous and in need of clarification. This post, the third in a series (see here and here) examining the impact of cyber on the law of naval warfare through the lens of the updated commentary to GCII, examines how the advent of cyber operations introduces an additional element of ambiguity.

Article 18 requires “[a]fter each engagement, Parties to the conflict shall, without delay, take all possible measures to search for and collect the shipwrecked, wounded and sick, to protect them against pillage and ill-treatment, to ensure their adequate care, and to search for the dead and prevent their being despoiled.” Para 1617 of the updated commentary to GCII rightly recognizes that “Article 18(1) is among the most important provisions in the Second Convention,” and that it sets out the obligations flowing from the protections accorded in Article 12. Therefore, a detailed understanding of each element in this article is key to a proper understanding of the entire convention. Here, we look solely at the first element in the light of cyber operations.

The first element makes plain that, unlike land operations, the requirement to tend to the sick and wounded does not arise until following the engagement. This is understandable in the naval context given the increased risk of harm a commander would endure by breaking off an engagement to collect the shipwrecked, wounded and sick. Keep in mind this obligation applies “without discriminating between their own and enemy personnel.” (Para 1618) Furthermore, at the time of GCII’s drafting, naval engagements tended to be very violent, but short-lived affairs. In the case of the Bismarck, the engagement was clearly ended when the ship, her ensign never struck, went under the sea after two hours of fighting. However, for modern navies equipped with advanced long-range weapon systems, including cyber capabilities, the end of the engagement may be more difficult to discern.

In the updated commentary, para 1648 discusses the article’s post-engagement limitation, stating that “unlike in land warfare, there is no requirement to undertake search and rescue activities during an engagement.” The commentary then argues that whereas this element may limit the obligation temporally, it may expand the obligation’s material scope. It reasons that “since the particular engagement will have ceased, this may limit the extent to which a Party to the conflict may invoke security or military considerations as a justification for not undertaking search and rescue activities.” Thus, determining the exact scope of the temporal requirement is vital.

Fortunately, the updated commentary provides guidance on interpreting the temporal clause of Article 18. Para 1655 provides that “the term ‘engagement’ is ‘a battle between armed forces’, i.e. involving the use of methods and means of warfare between military units of the Parties to the conflict.” Pre-empting the question of whether the methods and means are limited to the naval forces, the commentary suggests it “covers any kind of engagement, including from the air or from land but inflicting casualties at sea.” Cyber operations are not explicitly mentioned here, so it is worth discussing whether the cessation of cyber operations, in addition to the conclusion of more traditional kinetic operations, is required to “end the engagement” and initiate potential Article 18 obligations.

First, the commentary’s suggestion that “inflicting casualties at sea” is required for an engagement is most likely poorly worded. It is easy to imagine that ships may be engaged prior to actually inflicting casualties. Prior to her own sinking, the Bismarck sunk the HMS Hood in large part by achieving the “weather gage,” gaining an advantageous position in relation to the enemy prior to opening fire. Therefore, simply because a cyber-operation does not inflict casualties, this should not signal that operations is not part of the overall engagement.

Although the commentary to Article 18 does not refer to cyber operations, they are discussed in relation to the scope of application provisions of Article 2. Specifically, the question asked is whether cyber operations alone can constitute “armed force,” making the Geneva Conventions applicable. Para 277 states that “[i]t is generally accepted that cyber operations having similar effects to classic kinetic operations” would suffice. However, para 278 recognizes the current reality that cyber operations falling beneath this threshold are legally unsettled. It is safe to say that cyber operations achieving a kinetic effect, therefore, would continue the engagement. But what of those cyber operations that effect network systems without achieving kinetic effects?

Until such time as the jus in bello develops more fully in this area, it may be necessary to leave the legal reasoning to a good faith assessment by the ship’s commander. Although this seems initially unsatisfying, it is consistent with the new commentary’s understanding of Article 18. Para 1655 states that “[w]hat constitutes an engagement in any given case will remain context-specific,” and that “those acting on behalf of the Party to the conflict, each at his or her own level of decision-making, will need to make a good-faith assessment as to the moment it becomes possible to take one or more of the measures referred to in Article 18.” Such “good faith assessments” are a common and necessary part of IHL, even if open to occasional abuse.

Given the potential for abuse, what are nations employing cyber operations as part of naval conflicts to do? Parties to a conflict still have a vested interest in ensuring that the shipwrecked, sick, and wounded are recovered and cared for as quickly as possible. The commentary once again provides a potential solution. Para 1651 suggests that opposing commanders reach a “special agreement” on the rescue of those shipwrecked in the sense of Article 6, allowing parties to fulfill Article 18 obligations without fear of attack, adding that “such an agreement may be concluded orally, between commanders on the spot.” Alert commanders will be sure to add prohibitions on cyber-attacks as part of any such agreement.

A Potentially Serious Problem with the Final Decision Concerning Comoros

by Kevin Jon Heller

A couple of days ago, the OTP finally announced what we all expected: that it would not reconsider its refusal to open a formal investigation into Israel’s attack on the MV Mavi Marmara. Dov Jacobs has already offered some thoughts on the lengthy document the OTP has filed with the Court explaining its reasoning — what the OTP nicely calls the Final Decision. I fully concur with Dov’s thoughts (except with his position on retroactive acceptance of jurisdiction), and I write here simply to add one of my own.

To begin with, I think this is the most impressive OTP brief I have ever read — especially given the complexity of the procedural issues that it addresses. It is exceptionally well written and argued. I don’t know who the author is, but she would have made an excellent analytic philosopher. Fatou Bensouda should promote her immediately.

That said, I strongly believe that the Final Decision’s understanding of when the OTP is required to investigate a situation is fundamentally flawed — and will almost certainly come back to haunt the OTP in future preliminary examinations. I have argued, as have most scholars, that situational gravity is a function of all the potential cases in a situation that would be admissible before the Court: the greater the number of prosecutable crimes and the greater their individual gravity, the more situationally grave the situation. To be sure, it is not an easy task to compare the situational gravity of different situations. But I don’t think there a practical alternative, given that the OTP can only investigate a very small percentage of the situations in which admissible crimes have been committed.

The Final Decision, however, appears to take a very different approach. Instead of deciding whether to open an investigation based on the gravity of all the potentially admissible cases in a situation, the OTP seems to believe that it is required to open an investigation as long as even one potential case within a situation would be sufficiently grave to prosecute. Consider the following paragraphs (emphasis mine):

11. Although the Prosecution maintains its view that no potential case arising from this situation would be admissible before this Court—which is the only issue in dispute with the Comoros—this does not excuse any crimes which may have been perpetrated.

332. Consistent with article 53(3)(a) of the Statute and rule 108(3), and based on the above reasoning and the information available on 6 November 2014, the Prosecution hereby decides to uphold the disposition of the Report. There remains no reasonable basis to proceed with an investigation, since there is no reasonable basis to conclude that any potential case arising from the situation would be of sufficient gravity to be admissible before the Court.

This approach, it is worth noting, appears to represent a retreat from the position the OTP took in its initial explanation of why it would not investigate the Comoros situation. Here is paragraph 24 of that document (emphasis mine):

Having carefully assessed the relevant considerations, the Office has concluded that the potential case(s) that would likely arise from an investigation of the flotilla incident would not be of sufficient gravity to justify further action by the Court, in light of the criteria for admissibility 8 provided in article 17(1)(d) and the guidance outlined in article 8(1) of the Statute.

It is possible, of course, that the Final Decision refers to the gravity of “any potential case” instead of “the potential case(s)” not because the OTP’s approach to situational gravity has changed, but because there is only one potential case in the Comoros situation: the attack on the MV Mavi Marmara. But the difference of language is striking — and given the legal and analytic precision of the Final Decision, I find it difficult to believe that its emphasis on whether any individual case would be admissible is simply a slip of the keyboard.

I assume, therefore, that the Final Decision means what it says: the OTP believes it has to investigate any situation in which there is at least one potential case that is grave enough to be admissible. But that is a very problematic position.

To begin with, it leads to precisely the kind of unhelpful dispute we have seen in Comoros situation, where the OTP believes a specific case is not sufficiently grave to be admissible and the Pre-Trial Chamber disagrees. Both the OTP and the PTC have spent a great deal of time during their “judicial dialogue” (Dov’s apt expression) comparing the Mavi Marmara case to the Abu Garda and Banda cases. Here, for example, is how the Final Decision critiques the PTC’s insistence that the Mavi Marmara case is sufficiently grave to be admissible:

77. However, the Request does not address the basis on which the Prosecution considered that “the total number of victims of the flotilla incident reached relatively limited proportions as compared, generally, to other cases investigated by the Office”—in particular, the circumstances of the Abu Garda and Banda cases (which are, in relevant part, identical). Although the majority likewise referred to these cases, it did not consider those particular characteristics.

78. As the Report expressly states, Abu Garda likewise concerned the allegation of “a single attack involving a relatively low number of victims”—but it was “distinguishable” because of “the nature and impact of the alleged crimes”, which were committed against international peacekeeping forces. Accordingly, the attack alleged in Abu Garda differed in nature from the identified crimes aboard the Mavi Marmara. Crimes against international peacekeepers strike at the heart of the international community’s mechanisms for collective security, and thus their direct and indirect victims include not only the peacekeepers and their families, but also the large number of civilians deprived of protection more widely because of the disruption to the peacekeepers’ operations. The Request does not address this distinction. [130]

n. 130 Likewise, the recent Al Mahdi case—solely concerning attacks on property protected under article 8(2)(e)(iv) of the Statute—was considered sufficiently grave to be admissible before the Court, resulting in a conviction. In the context of sentencing, the Trial Chamber stressed that the charged conduct was of “significant gravity”, among other reasons, because 1) the destroyed mausoleums were “among the most cherished buildings” in Timbuktu, an “emblematic city” which “played a crucial role in the expansion of Islam in the region” and which is “at the heart of Mali’s cultural heritage”; 2) the destroyed mausoleums were of proven significance to the inhabitants of Timbuktu not only as a matter of religious observance but also as a symbol and focus of community activity and unity; and 3) all the destroyed sites but one were designated UNESCO World Heritage sites, whose destruction also directly affects “people throughout Mali and the international community.” This same reasoning is applicable, mutatis mutandis, to the question of admissibility.

I don’t find the OTP’s efforts to distinguish the Mavi Marmara case from Abu Garda, Banda, and Al Mahdi particularly convincing. Its selection of factors to highlight strikes me as completely subjective and result-driven. Indeed, when faced with the PTC’s insistence that the message the Mavi Marmara attack sent to the international community — that Israel is willing to use force to maintain an illegal blockade that is causing a massive humanitarian crisis in Gaza — it simply retreats to “well, we disagree, and there is nothing you can do about it”:

80. Indeed, the majority appears simply to disagree with the Prosecution’s view of the weight to be given to… the significance of any ‘message’ sent by the interception of the flotilla itself. Given the Prosecution’s understanding of the proper standard of review under article 53(3)(a), and the absence of a reasoned conclusion that the Report was in these respects incorrect or unreasonable, the Prosecution does not consider it appropriate to depart from its original determination in the Report.

My point is not that the PTC’s gravity analysis is right and the OTP’s is wrong. (Though I do think the PTC has the stronger argument.) My problem is with the OTP’s position that it must investigate any situation in which at least one case is grave enough to be admissible. Debates over case gravity are inevitable when that is the standard for opening an investigation. But they are easily avoided if the OTP takes a more holistic approach to situational gravity, comparing the gravity of different situations by examining all of the potentially admissible cases within them. Even if we assume (as I do) that the attack on the Mavi Marmara is sufficiently grave to be admissible, the overall situational gravity of the Comoros situation (which involves only one case) still pales in comparison not only to numerous other situations under preliminary examination, but even — and more importantly — to the situational gravity of the Palestine situation as a whole. As I have argued previously, the last thing the OTP should do is investigate one very small part of the much larger conflict between Israel and Palestine. If it ever takes the Palestine situation on, it needs to look at crimes committed by both sides throughout Palestinian territory.

There is, however, an even more significant problem with the Final Decision’s standard for opening an investigation: if taken seriously, it will simply overwhelm the OTP’s resources. There may not be even one admissible case in the Comoros situation (because there is only one case), but how likely is it that larger situations, which are the norm, will not contain even one case sufficiently grave to prosecute? Just think about the situations currently at Phase 2 or Phase 3 of the preliminary-examination process: Burundi, Gabon, Iraq, Palestine, Ukraine, Colombia, Guinea, and Nigeria. There may well be complementarity issues in some of those situations that counsel not opening an investigation, but it seems exceptionally likely that each contains at least one admissible case. The Final Decision’s standard would thus seem — barring complementarity concerns — to require the OTP to open a formal investigation in all eight situations. Which is, of course, practically impossible.

Nor is that all. If the existence of even one admissible case is enough to require the OTP to investigate a situation, states will have little problem using referrals (self or other) to achieve nakedly partisan ends. Palestine, for example, could simply refer a single day during Operation Protective Edge in which Israel flattened an entire neighbourhood in Gaza or destroyed a UN school sheltering displaced civilians. It would be difficult, if not impossible, for the OTP to plausibly maintain that those acts are not grave enough to prosecute. So it would have to open an investigation. That makes little sense. Far better for the OTP to simply say that, however grave those specific attacks might be, the overall gravity of the gerrymandered “situation” is not sufficient to investigate in light of the gravity of other situations.

I hope I am wrong about when the OTP believes it is required to open an investigation into a situation. If so, the OTP needs to clarify its position immediately. Because the standard articulated in the Final Decision — the existence of even one case sufficiently grave to be admissible — is simply unworkable.