Torture and intelligence sharing is back in the news, again. Last month, it was in-bound intelligence -- that is intelligence from torturing states. I offered my view about the government's policy on this issue here. This week, it is out-bound intelligence sharing; the sharing of intelligence with torturing states. The key issue: what to do when that intelligence-sharing may be causally related to torture by a torturing state. Once again, the key reporting on this comes from Jim Bronskill.
The government's answer: well, subject to the weighing of a number of considerations, it may still be shared. And in so opining, I believe that the government is offside the law.
The Ministerial Directive
The directive commences with a brief (and, with respect, imperfect) summary of legal principles. More on that below. It then discusses the concept of outbound intelligence sharing with torturing states in circumstances a) where there is no substantial risk of mistreatment and b) when there is substantial risk of mistreatment. The document defines "substantial risk" and sets up differential approval mechanisms for the information sharing depending on whether that substantial risk exists or not.
The "substantial risk" issue doesn't quite come from the ether -- it is the concept housed in Art. 3 of the Torture Convention when deciding whether someone can be deported to a torturing state. (Although there it is called "substantial grounds"). The analogy to outbound information sharing may be imperfect, but I understand where it comes from.
So while one could muse at length about how this substantial risk is defined, I shall focus instead on what happens where CSIS concludes there is a "substantial risk" that information sharing will induce torture. The text reads:
When there is a substantial risk that sending information to, or soliciting information from, a foreign entity would result in the mistreatment of an individual, and it is unclear whether that risk can be mitigated through the use of caveats or assurances, the matter will be referred to the Director [of CSIS] for decision.
In making his or her decision, the Director will normally consider the following information, all of which must be properly characterized in terms of its accuracy and reliability:
• the threat to Canada's national security or other interests, and the nature and imminence of that threat;
• the importance of sharing the information, having regard to Canada's national security or other interests;
• the status of the relationship with the foreign entity with which the information is to be shared, and an assessment of the human rights record of the foreign entity;
• the rationale for believing that there is a substantial risk that sharing the information would lead to the mistreatment of an individual;
• the proposed measures to mitigate the risk, and the likelihood that these measures will be successful (including, for example, the foreign entity's record in complying with past assurances, and the capacity of those government officials to fulfil the proposed assurance);
• the views of the Department of Foreign Affairs and International Trade (DFAIT); and
• the views of other departments and agencies, as appropriate, as well as any other relevant facts that may arise in the circumstances.
The Director may refer the decision whether or not to share information with the foreign entity to the Minister of Public Safety, in which case the Minister will be provided with the information described above.
The Director or Minister of Public Safety shall authorize the sharing of information with the foreign entity only in accordance with this Direction and with Canada's legal obligations.
A Hierarchy of Values
This policy says really too things. First, the considerations it lists suggests that where the objective of the sharing is important enough, information can be shared. Second, it says that where the substantial risk can be mitigated, information can be shared. There is no hierarchy here. It does not say that the two must go together. And so mitigation is not a pre-requisite. It is a consideration. Indeed, if the torture risk can be mitigated, then presumably there is no substantial risk of torture anyway.
On mitigation, enter the lengthy and thorny discussion of so-called "assurances" -- promises by torturing countries to not do things their law codes say are alreay illegal. My view (and my doubt) about assurances can be found here.
The even more disturbing issue is the one about sharing intelligence if the stakes are high enough, even if it means torture. This is all contrary to the Arar commission report, of course. That report acknowledged the importance of information sharing but said there was an outer limit. But even more critically, the government cannot do this information sharing and still, with a straight face, also claim that it will "authorize the sharing of information with the foreign entity only in accordance with this Direction and with Canada's legal obligations."
Canada's Legal Obligations
In the aforementioned summary of Canada's legal obligations, the government forgets one key provision of the Torture Convention. It reads: "Each State Party shall ensure that all acts of torture are offences under its criminal law. The same shall apply to an attempt to commit torture and to an act by any person which constitutes complicity or participation in torture" (Art. 4, emphasis added). By the way, the prohibition on torture is a jus cogens norm. There is an entire essay to write on this, but insofar as I am concerned, that means that Canada's other treaty obligations (for instance, honouring Security Council resolutions on information sharing in anti-terrorism) are subordinated to this obligation.
Let's assume that sharing information that induces torture doesn't amount to "participation" (and actually, I'm not sure I would be quick to bet on this). The question then becomes: "what is complicity"?
"Complicity" isn't really a robust legal term of art. That is, it is an almost colloquial expression that captures (but is not limited) to better understood legal concepts such as aiding and abetting. So for lack of a better starting point, how far can aiding and abetting reach? In international law, very far indeed.
The International Criminal Court for the Former Yugoslavia has treated the aiding and abetting as part of customary international criminal law and has described its elements as follows:
(i) The aider and abettor carries out acts specifically directed to assist, encourage or lend moral support to the perpetration of a certain specific crime (murder, extermination , rape, torture, wanton destruction of civilian property, etc.), and this support has a substantial effect upon the perpetration of the crime. ...
(ii) In the case of aiding and abetting, the requisite mental element is knowledge that the acts performed by the aider and abettor assist [in] the commission of the specific crime of the principal. ...
There need not be a cause-effect relationship between the assistance provided, and the ultimate crime. At issue, simply, is whether the assistance has a substantial effect on the occurrence of the crime. The ICTY has concluded, for instance, that detentions commanded by one accused assisted in persecutions then committed by other accused, and constituted aiding and abetting. In another instance, the ICTY found an accused guilty of aiding and abetting where he “was aware that the prisoners were being mistreated by … soldiers [over which he had no control] on a recurring basis over a period of time (without specifying the precise nature of that mistreatment), yet with that awareness he continued to participate in sending the prisoners out to work under those soldiers and (having responsibility for the welfare of the prisoners) he failed to take measures open to him to stop them from going out to work in such conditions.”
Notably, “knowledge on the part of the aider and abettor that his acts assist in the commission of the principal perpetrator's crime suffices for the mens rea requirement of this mode of participation.” It “is not necessary that the aider and abettor...know the precise crime that was intended and which in the event was committed. If he is aware that one of a number of crimes will probably be committed, and one of those crimes is in fact committed, he has intended to facilitate the commission of that crime, and is guilty as an aider and abettor.”
Apply this to the CSIS outbound intelligence context. Knowing full well of the substantial risk that torture will result from the sharing of the information, CSIS nevertheless shares the information. This then has a "substantial effect" on the occurrence of the crime -- by definition in fact. (If CSIS says, "Pick up Joe. He's a terrorist." And the torturing states acts on the tip and then tortures, it's really hard to see how that the tip isn't assistance that has a substantial effect). And there is absolutely foreknowledge. The elements of the crime seem to be satisfied. Indeed, the ministerial directive is almost a checklist of those elements.
Now, a couple of caveats. I am drawing on ICTY jurisprudence on the content of customary international law in the area of war crimes to inform the content of the word "complicity" in the Torture Convention. It is an extrapolation necessitated by the sparse jurisprudence in international law, but not a foregone conclusion. It is sure one I would be prepared to offer up, though, if I were giving legal advice on this. More than that, I am very certain that the UN Committee Against Torture, presented with the ministerial directive, will not pause in concluding it constitutes complicity (for what such a CAT view is worth -- not much really, given Canada's past indifference in other instances).
Second caveat: the ICTY concept of aiding and abetting is broader than the Canadian equivalent. Whether an action aided and abetted torture under the Criminal Code would be decided by reference to s. 21 of that statute. The latter reads: “Every one is a party to an offence who (a) actually commits it; (b) does or omits to do anything for the purpose of aiding any person to commit it; or (c) abets any person in committing it.” The actus reus of aiding or abetting is: “any act or omission that occurs before or during the commission of the crime, and which somehow and to some extent furthers, facilitates, promotes, assists or encourages the perpetrator in the commission of the crime…, irrespective of any causative role in the commission of the crime.”
In terms of the mental element, the reference to “purpose” could suggest that the accused desire the ultimate outcome. This is, however, not the approach adopted by the Supreme Court of Canada. Instead, “purpose” is equated with “intent”, and “does not require that the accused actively view the commission of the offence he or she is aiding as desirable in and of itself”. The Ontario Court of Appeal has subsequently held that the normal meaning of the words “purpose” and “intend” “suggests that a person must subjectively advert to a specific objective and that he or she, therefore, must have knowledge of the facts that constitute that objective”. But more than that, the accused must intend that his or her actions aid or encourage the perpetrator: the assistance must be provided “with the intention of helping the (or, a) principal to commit the offence.” Mere knowledge of the perpetrator’s criminal acts or propensities may prompt a legitimate inference that the accused intended to assist, but knowledge alone is not enough: “while knowledge can found an inference of intention, it alone cannot constitute the requisite mens rea”.
In other words, here, the CSIS intelligence sharing would have to be done with more than knowledge of torture, but also with the intention of helping the torturer torture. That is, intend to induce the torture. The ministerial policy is geared towards making the torture sound like the unfortunate consequence of the torture. I do not read it as authorizing CSIS to release information for the purpose of inducing torture. If it is read that way, people at CSIS had best consult legal counsel.
The Charter and All That
The Canadian Charter of Rights and Freedoms stands in the way of actions in Canada that visit harm overseas, even when that harm is visited on the victim by a foreign government. That is the lesson of Burns. One of these harms is torture. That is the lesson of Suresh. The only difference between those two cases and the present discussion is that Burns and Suresh were in Canada and Canada was choosing whether to remove them to that foreign harm. Here, the victim does not necessarily have a nexus to Canada. But the causal chain leading to the harm still commences here.
I have no doubt that the government is distinguishing this policy from Suresh on the nexus issue, and the perplexing Federal Court of Appeal jurisprudence in the Amnesty International Afghan detainnee cases that seems preoccupied with the citizenship of the victim when it comes to the extraterritorial reach of the Charter, seemingly extracted from a narrow read of the facts in Khadr. But that is slender grounds on which to base one's constitutional hopes. I simply do not think that the Hape and Khadr line of cases from the Supreme Court stand for the principle that the Charter does not apply overseas when there is a insufficient nationality nexus. Instead, the issue is whether Canada's international obligations apply to that context. And on that test, and for the reasons outlined above, they do in this torture context.
All of this is to say that this latest ministerial directive drinks deeply of the dark side. I await the next commission of inquiry and some bountiful years of litigation.
Domestic Aiding and Abetting Notes
 R. v. Dooley, 2009 ONCA 910 at para. 123.
 R. v. Hibbert,  2 S.C.R. 973 at paras. 31 and 39.
 R v. Helsdon, (2007) 84 OR (3d) 544 (On CA) at para. 28. The Court of Appeal treated aiding and abetting identically, despite the fact only “aiding” includes a reference to “purpose”. See also R. v. Almarales, 2008 ONCA 692 (doing the same).
 R. v. Almarales, 2008 ONCA 692 at para. 66.
 R. v. Palombi, 2007 ONCA 486 at paras. 18 and 19. After this paper was prepared, the Supreme Court of Canada affirmed that the mens rea of aiding and abetting includes both knowledge and intent. R. v. Briscoe, 2010 SCC 13.
 See, e.g., Prosecutor v. Furundzija, IT-95-17/1, Judgment (10-12-1998) (ICTY) at para. 191 et seq.
 Prosecutor v. Vasiljevic, IT-98-32-A, Appeal Chamber Judgment (25-02-2004) (ICTY) at para. 102.
 Prosecutor v. Blagojevic, IT-02-60-T, Judgment (17-01-2005) (ICTY) at para. 733 et seq.
 Prosecutor v. Aleksovski, IT-95-14/1-A, Appeal Chamber Judgment (24-03-2000) (ICTY), at paras. 170 and 172.
 Prosecutor v. Blaskic, IT-95-14-A, Appeal Chamber Judgment (29-07-2004) (ICTY), at paras. 49 and 50.