What to do with tortured intelligence: Room for an 'operational' vs legal use distinction?

A recurring controversy in Canada is the treatment by the Canadian security services of intelligence suspected of being the product of torture, supplied by allied agencies. 

In its most recent annual report, the Security Intelligence Review Committee recorded its belief "that information-sharing with countries that have poor human rights records will continue to be a difficulty for CSIS until the Government of Canada resolves its seemingly contradictory
position on information obtained from torture, versus the directions it
provides to CSIS on carrying out its work."

In its own annual report, CSIS discusses its information-sharing and human rights policies, although its clear focus is on "out-bound" information, that is information provided by Canada to the foreign partner. 

The suspect provenance of "in-bound" information -- information provided by the foreign partner -- has been an issue in various immigration security certificate cases and, more recently, in parliamentary discussions, including on CSIS's role in Afghanistan.

It is worth contemplating questions of both legality and legitimacy stemming from these latest discussions on use of in-bound torture intelligence.

Two Snapshots on Use of Tortured Intelligence

In March 2009, a CSIS lawyer sparked controversy in testimony before a parliamentary committee with the following statement:  "...do we use information that comes from torture [supplied by foreign agencies]? And the answer is that we only do so if lives are at stake."

More recently, on May 5, 2010, a second CSIS witness told another parliamentary committee that it was possible that some information supplied by Afghan security services to CSIS was the product of torture, which should result in the information then being flagged and corroborating evidence sought. 

These comments generated controversy and commentary.  But it would be unfair to condemn outright these views without some careful review.  The issue of what to do with in-bound information that may be the product of torture is actually a fairly complex one.

Clear Rules

Here are the simple rules:

1. Torture evidence is illegal: It would violate both domestic law (Criminal Code, s.269.1(4) and in the security certificate context, Immigration and Refugee Protection Act, s.83(1.1)) and public international law (UN Torture Convention) to admit as evidence information procured via torture (and in the case of IRPA, cruel, inhuman or degrading treatment).  Efforts to use this information in court (or administrative) proceedings would also undoubtedly violate s.7 of the Charter of Rights and Freedoms and common law rules governing the admissibility of evidence, where these concepts apply.  Fuller discussion on all of these issues is found in National Security Law, ch. 15.

2. Complicity with torture is illegal: Under Art. 4 of the UN Torture Convention, states are to outlaw "complicity" with torture.  Under s.7(3.7) of the Criminal Code, it is a crime, in relation to torture taking place overseas, to engage in

“a conspiracy or an attempt
to commit an offence against, being an accessory after the fact in relation to
an offence against, or any counselling in relation” to torture.  Aiding and abetting torture also renders the accused a party to the crime (although the question of whether extraterritorial aiding and abetting is criminalized under the Criminal Code may be a live one).  (Aiding and abetting torture in a non-international conflict like that in Afghanistan would, however, likely be prosecutable under the Crimes Against Humanity and War Crimes Act, as a war crime.  For a detailed dicussion of this, see this report).  All told, it seems likely that "complicity" (of various sorts) with torture is criminalized in Canadian criminal law.

Absolutism or Pragmatism?

The issues become more complex past this point.  Is it wrong to rely on information that may be the product of torture in other circumstances?  For instance, could one use the information only for "operational" purposes?  One answer is that any reliance (of whatever sort) on torture amounts to tacit acceptance of the practice and is morally wrong (even when it is not technically illegal).  The International Commission of Jurists' Eminent Panel on Terrorism, Counter-Terrorism and Human Rights put it this way:

Some States accept the prohibition on intelligence gathered under torture being used in legal proceedings, but at the same time seem ready to justify the use of such intelligence for operational purposes (for example to prevent a terrorist attack). The first usage is supposedly more acceptable (or at least less reprehensible) than the latter.

This differentiation between the use of information obtained by torture and other cruel, inhuman or degrading treatment, for “legal” and for “operational” purposes is problematic for several reasons. It undermines the absolute prohibition on torture which entails a continuum of obligations – not to torture, not to acquiesce in torture, and not to validate the results of torture and other cruel, inhuman or degrading treatment. Secondly, it suggests a water-tight distinction between “legal” and “operational” use which is probably illusory, and certainly the Panel was supplied with examples where information was supposedly sought on operational grounds, but subsequently relied upon in legal proceedings that followed. Thirdly, States have publicly claimed that they are entitled to rely on information that has been derived from the illegal practices of others; in so doing they become “consumers” of torture and implicitly legitimise, and indeed encourage, such practices by creating a “market” for the resultant intelligence. In the language of criminal law, States are “aiding and abetting” serious human rights violations by others.

Likewise, the UN special rapporteur on human rights and counter-terrorism urges that "States must not aid or assist in the commission of acts of torture, or recognize such practices as lawful, including by relying on intelligence information obtained through torture. States must introduce safeguards preventing intelligence agencies from making use of such intelligence."  The rapporteur cites Justice Neuberger, dissenting in A and Others v. Secretary of State for the Home Department, para. 497: “(…) even by adopting the fruits of torture, a democratic State is weakening its case against terrorists, by adopting their methods, thereby losing the moral high ground an open democratic society enjoys”.

Of course A and Others concerned use of tortured evidence in court proceedings, an issue discussed above as one where the rules are relatively straightforward.

A and Others did not deal with more difficult situations where there is less clear law.  For instance, if CSIS were to receive a report from a notorious (but often reliable) foreign security service that a bomb would be detonated on the Toronto subway system in 30 minutes  (suspecting or even outright knowing that the information in that report was elicited by torture), should it turn a blind eye to that information?  Imagine the consequences of doing so, if it then turns out that the information was accurate (and while I am a strong proponent of the view critiquing the reliability of torture information, I am prepared to admit that sometimes it could turn out to be accurate).  A bomb goes off, people die and the security service did nothing.  The fall out -- including as measured in terms of blood, treasure and civil liberties -- would be dramatic and tragic.


In this hypothetical, the only reasonable response by the authorities is, in my view, to take immediate precautions on the Toronto subway system.  Because of those precautions, there has been "use" or "reliance" on the tortured intelligence, but of a very different sort than, for example, using it in a security certificate proceeding.  This reliance in the subway scenario does not itself impair (in any meaningful way) the rights or liberties of a person on the basis of tortured evidence. 

In these circumstances, and given the "opportunity cost" of not using the information (possible death, injury and political fall-out from these events), the absolutist view on non-use comes close to fiat justitia et pereat mundus ("let justice be done, though the world perish").

I suspect that this is the sort of scenario the CSIS witness had in mind in March 2009 when he told parliamentarians that the Service would use torture intelligence if "lives are at stake".

It is also a scenario addressed, implicitly, by the Ottawa Principles on Anti-terrorism and Human Rights.  That instrument -- the product of an academic conference in 2006 -- takes what I regard as the pragmatic approach to this question, rather than outright barring use of information procured by torture in every conceivable situation:

Information, data, or intelligence that has been obtained through torture or cruel, inhuman or degrading treatment or punishment may not be used as a basis for

(a) the deprivation of liberty;
(b) the transfer, through any means, of an individual from the custody of one state to another;
(c) the designation of an individual as a person of interest, a security threat or a terrorist or by any other description purporting to link that individual to terrorist activities; or
(d) the deprivation of any other internationally protected human right.

To be clear, principle 4.3.2. is not a statement of law but rather a way of measuring legitimacy.  It addresses at least one of the International Commission of Jurists' criticisms about the distinction between "operational" and "legal" use of the information.  It would also allow the government to react in the Toronto subway scenario.  

Principle 4.3.2. is, in my view, an appropriate compromise position on the question of use of torture intelligence.

The Afghanistan Problem

A final issue is whether the second "snapshot" on use of torture information discussed above -- the consumption of information by Canada that might have been procured by torture by Afghan authorities -- is legal or legitimate. 

According to the most recent CSIS annual report, "[i]n 2008-09, CSIS also continued to provide security intelligence
support to the Canadian Forces in Afghanistan. CSIS has had a presence
in the country for the past few years and also continues to gather
intelligence in the region in order to mitigate potential security
threats to Canada which have a nexus to Afghanistan."

In testimony before the parliamentary committee on May 5, the CSIS witness observed that this Afghan activity by CSIS has led to the disruption of insurgency networks engaged in "imminent" bombing attacks against civilians and military forces.  The CSIS witness also noted that: 1. Canada did not transfer detainees to Afghan authorities for the purposes of extracting information from detainees; 2. It was not certain whether any torture intelligence stemmed from any detainee transferred by Canada to Afghanistan (pursuant to the Canadian Forces transfer policy); and 3. It may be that some of the information received from Afghan authorities was procured by torture, in which case, under CSIS policy, that information should have been "caveated" and could not be relied upon exclusively without corroborating information.

Focusing on point 3, it is of course the case that caveating and verification of tortured intelligence would not forgive complicity in torture, discussed above.  If, for example, Canadian officials counseled or aided and abetted torture, that would still be criminal, with all the caveating and verification in the world.  On the other hand, passive consumption of information for intelligence purposes that does not amount to, e.g., counseling or aiding and abetting is likely not illegal. 

The question is whether such practice is, nevertheless, still legitimate.  The absolutist view would be that, even with caveating and verification, the consumption of the fruit of torture has the effect of creating a market for it, and tacitly endorsing it.  (The International
Commission of Jurists' comment about serving as a market for this
information itself, alone amounting to "aiding and abetting" would not
be true as a matter of Canadian criminal law).

There is more merit for this absolutist view in the Afghan context than in the Toronto subway hypothetical above.  In my Toronto hypothetical, the information appears, and the government decides whether to act.  In the Afghan context, the relationship appears to be more structured; that is, there would presumably be a degree of organized flow of information, and consumption of it, that in both quantity and frequency exceeds the more ad hoc scenario described in my subway hypothetical.  Put another way, the marketplace for the torture intelligence created by Canada is a fairly significant one.


It is also the case that because Canada then acts on that information to, e.g., disrupt insurgency networks, that means that someone's right and liberties are being affected in a manner much greater than in the subway hypothetical, perhaps in a manner that exceeds principle 4.3.2 of the Ottawa Principles.  

Here, however, the caveating and the verification could be important: if it operates properly, with caveating and verification, the torture intelligence may spark further (presumptively lawful) intelligence collection, but only after that further intelligence collection are actions then taken that affect rights and liberties of the sort enumerated in principle 4.3.2..  In those circumstances, principle 4.3.2 is likely appeased.  That, in my view, would strongly enhance the legitimacy of the intelligence practices in Afghanistan.  But before arriving at a conclusion on this, it would certainly be very helpful to know more about the CSIS caveating and verification policies and practices.