The Book





This blog comments on Canadian (and occasionally comparative) national security law to update my National Security Law textbook and now also my 2015 book, False Security: The Radicalization of Anti-terrorism, co-authored with Kent Roach.

Please also see for Bill C-51-related analyses by Craig Forcese and Kent Roach.

For narrated lectures on various topics in national security law, please visit my 2017 "national security nutshell" series, available through iTunes.

Please also visit my archive of "secret law" in the security area.

By Craig Forcese

Full Professor
Faculty of Law

Email: cforcese[at]

Twitter: @cforcese


National Security Law Blog Search
Subscribe to National Security Law Blog

Best Law School/
Law Professor Blog Award


Most Recent Blog Postings

A Three Front Legal Campaign: CSEC, Metadata and Civil Liberties

There are now three lawsuits in Canadian courts related to the Communications Security Establishment Canada's intercept of metadata.  These are:

  • BC Civil Liberties Association's action for a declaration in BC Superior Court alleging that CSEC's intercept activities violate sections 8 and 2 of the Charter.
  • BCCLA's class action in Federal Court, brought "in order for those persons whose private communications and/or metadata have been intercepted to have access to remedies under s.24 of the Charter once the issues of the unconstitutionality of the impugned provisions, Authorizations and Directives are resolved in the [BC Supreme Court] Declaratory Action".
  • Canadian Civil Liberties Association (CCLA)'s application for a declaration in Ontario Superior Court that sections 7 and 8 of the Charter are violated by those provisions in the Federal Personal Information and Protection of Electronic Documents Act (PIPEDA) that permit government agencies (including, it is alleged, CSEC) to obtain personal information from Canadian telecommunications companies with prior judicial authorization.

These cases overlap.  The BCCLA class action is intended to be derivative of its BC Superior Court declaratory action.  The CCLA case seems likely to engage the question of a person's reasonable expectation of privacy of metadata in the possession of third party telecommunications providers.  That question must inevitably also arise in the BCCLA declaratory action.

Meanwhile, the Supreme Court of Canada will decide the Spencer case at any time.  This matter flows from a  Saskatchewan Court of Appeal decision where the police identified the accused by asking a telecom provider to provide the identifying information associated with an IP address.  There, a badly fractured appeal court proposed no clear rule on whether disclosure of this information violated a reasonable expectation of privacy or, if so, whether the PIPEDA provisions (that the police relied upon) were constitutional. 

In deciding the case, the Supreme Court will likely pronounce on these issues, with second order, knock-on effects for the three metadata related lawsuits noted above.  However, it is entirely possible -- and indeed probable -- that whatever they say on metadata here will not be the last pronouncement on the issue.  As I discuss in a draft article I recently posted on the issue, the metadata appellate cases to date have often been focused on child pornography police investigations.  Spencer falls into this category.  In deciding these cases, the Courts have generally been careful not to over generalize the constitutional issues before them, or to lump all "metadata" together.

For instance, in R. v. Ward, the Ontario Court of Appeal decided that search of subscriber information stripped the accused of “his Internet anonymity” and had the potential “to reveal activities of a personal and private nature”. Even so, and even with a strong subjective expectation of privacy, the Court of Appeal doubted the objective reasonableness of the privacy expectation.  However, the Court of Appeal also issued a caution:

the conclusion in this case is based on the specific circumstances revealed by this record and is not intended to suggest that disclosure of customer information by an ISP can never infringe the customer's reasonable expectation of privacy. If, for example, the ISP disclosed more detailed information, or made the disclosure in relation to an investigation of an offence in which the service was not directly implicated, the reasonable expectation of privacy analysis might yield a different result. Similarly, if there was evidence that the police, armed with the subscriber's name and address, could actually form a detailed picture of the subscriber's Internet usage, a court might well find that the subscriber had a reasonable expectation of privacy.

In my article, I argue that Ward suggests that metadata may, in fact, attract section 8 protections, and it implies that this likelihood increases in proportion to the sweep of the disclosure and the intimacy of the portrait that might then be painted from the disclosed information. Moreover, what is reasonable disclosure by an ISP in one instance might not be so reasonable in another. While ISPs may be equally compliant in practice, it does not follow that a court would conclude that acquiesce in a broad, search-of-the-haystack foreign surveillance effort is as reasonable as cooperation in a targeted child pornography police investigation in which the ISP’s services are used as a vessel for the crime.  In other words, the average user’s believe that their metadata are not subject to intelligence trolling via cooperative ISPs may be objectively reasonable. 

(If trolling is what, in fact, is happening.  So much of the legal discussion on CSEC is built on factual extrapolations, interpolations and plain speculation.  If a global driftnet sweeps up Canadian data by happenstance rather than by intent, does that matter for constitutional purposes?  I am inclined to think not.)

All of this is to say that in the world of search and seizure rules, what is good for the law enforcement goose may not be good for the intelligence service's gander.


Harkat Supreme Court Decision: Winners and Losers


The Supreme Court released its long anticipated decision in the Harkat security certificate case this morning.  I won't summarize the holding -- I can't do any better than the headnote writers at the Court.  But I think it worth reviewing the case from a "winners" and "losers" perspective, since the immediate buzz is that this is a win for the government.

The answer on this point is probably: "yes" this is a government win from a "headline" perspective, and "no" from a "we like what the Court has said about security certificates" perspective.

Mr. Harkat

This is clearly a huge loss for Mr. Harkat.  To be clear: That doesn't mean it's the end of his legal battles.  The issue may now move to whether he can, in fact, be removed from Canada.  If there is a risk of maltreatment upon removal, we will be back to a discussion about what the SCC meant in Suresh in relation to removal in the face of torture, and if its obiter statement suggesting that this was possible still reflects the current state of the law. (Personally, I don't think it does, since the Court has since re-acknowledged that the Charter should be construed consistently with Canada's international obligations, and there is no way Canada's international obligations permit removal to torture.)  A related issue is whether the decision on whether removal will result in maltreatment must be made by an independent decision-maker, as opposed to someone in the bureaucratic hierarchy.

But for security certificate purposes, the battle is over.  (And on this point: it must be particularly galling to Mr. Harkat and his legal team that the Court acknowledged a Charter violation in relation to the destruction by CSIS of the original transcripts of intercepted communications, but refused to do anything about it. To use such strong language to condemn the government action, and then to fail to exclude the intercept summaries as evidence and to send the matter back to be reconsidered on the remaining facts -- as the Federal Court of Appeal did -- is one of those SCC decisions I shall never be able to explain to my law students.  At some level, the Court decides to do this factual reassessment itself, serving as a proxy fact-finder, while talking about not being a fact-finder in the portion on Khattab discussed next).

And so at the end of the day, this turned out to be a case about facts not law.  And one of the most critical facts was "who was Ibn Khattab?".  Justice Noël, in Harkat, considered affiliation with Ibn Khattab justified the government's case because Khattab was connected to terrorism and "the Bin Laden network" (at para. 4).  In Justice Noël's words: "he is a member of the Bin Laden Network through his past work for the Khattab group" (as well as other associations) (at para. 548).

But just to show that these facts can be mutable things, Justice Mosley, in Almrei, concluded that Khattab was not part of Al Qaeda and that "association with Khattab does not, in my opinion, support a finding that he is a danger to the security of Canada" (at para. 464).

This differing view of the role and activities of one man was not enough to persuade the Supreme Court to vacate this case: "Noël J. was entitled to make his own assessment of whether Ibn Khattab was involved in terrorist activities, based on evidence that he found to be reliable and appropriate. I would not interfere with his assessment" (at para 109).

On such things, the fate of individuals turns.  Pretty messy.


The Government

But that Mr. Harkat lost does not mean that the government "won" and had its way, at least from the perspective of those of us interested in the institutional development of the special advocate system.  The Court didn't find reason to doubt the constitutionality of the system, because possible constitutional deficiencies can be preempted by following its close instructions on a number construals of the Act.  To list those:

  • Minimum Disclosure: There is an "incompressible minimum amount of disclosure to the named person".  To meet this standard, the named person must "personally [have] received sufficient disclosure to be able to give meaningful instructions to his public counsel and meaningful guidance and information to his special advocates which will allow them to challenge the information and evidence relied upon by the Minister in the closed hearings" (at para. 56).  "If the named person is not reasonably informed, the proceedings will not have been in compliance with the IRPA scheme and the judge cannot confirm the certificate’s reasonableness. In such a case, the judge must quash the certificate, pursuant to s. 78 of the IRPA" (at para. 60).  This is potentially a big deal, and probably not something the government is content to see.
  • No Failures to Disclose Based on Doubtful Hypotheticals or Disclosure Allergies:  "Only information and evidence that raises a serious risk of injury to national security or danger to the safety of a person can be withheld from the named person" (at para. 61). " The judge must be vigilant and skeptical with respect to the Minister’s claims of confidentiality" (at para. 63).  Again, not a great rule from the government's perspective.
  • Continued Communication is Real and to Be Preferred: "The judge should take a liberal approach in authorizing communications [between special advocates and named persons after the former has seen the secret evidence] and only refuse authorization where the Minister has demonstrated, on a balance of probabilities, a real — as opposed to a speculative — risk of injurious disclosure. As much as possible, the special advocates should be allowed to investigate the case and develop their strategy by communicating with the named person, the named person’s public counsel, and third parties who may bring relevant insights and information" (at para. 70).  And special advocates can make the case for continued communication ex parte, without the government lawyers present (and thus learning lots about the other side's litigation strategy).  Given where the government started on the communication back in the day and traditional position in places like the UK -- communication verboten -- this is a significant holding.  I haven't tracked how much communication has existed in practice, but this language leans towards "a fair amount" and possibly more than has heretofore been on offer.
  • There is no "CSIS informer privilege": "[T]he IRPA scheme already affords broad protection to human sources by precluding the public disclosure of information that would injure national security or endanger a person" (at para. 87).  Again, not the answer the government wanted.

Special Advocates

Special advocates win, to the extent you can view them as an actor in this matter.  They will continue, at least as a concept.  They will have more continued communication and in the right circumstances, they may get to cross-exam intelligence sources, although the Court was pretty restrained on this point. 

Other Named Persons

What impact this decision will have on the Jaballah (still pending) and Majhoub cases (being appealed?) remains to be seen.  But my strong suspicion (given the above) is that the SCC decision in Harkat has now moved the ball even further down the field from where the government would like it to be. Now after 7 years of effort we have a system with reasonably clear rules.  And since the government may not like where those rules take it, it is unlikely ever to issue another terrorism-related security certificate (if it can avoid it). 


Parliamentary Review of Intelligence Service Activity: Assessing the Different Models

As discussed in NSL, ch. 3, the Canadian Parliament’s record on national security matters compares unfavourably with that of other Westminster democracies. (For a more recent review of the Canadian Parliament and national security matters, see Nicholas MacDonald's interesting article). The United Kingdom, Australia and New Zealand have national security parliamentary committees, although they vary in their rigour.  Moreover, many (if not most) European states also use a parliamentary review committee model.

In 2004, the Canadian government tabled a discussion paper noting this comparative experience and identifying means of enhancing the parliamentary role in national security matters. That more prominent role was endorsed in the subsequent national security policy, which proposed the creation of a “National Security Committee of Parliamentarians.” An interim committee of parliamentarians on national security also provided their views in October 2004, recommending a statutorily created committee of Parliament.

Subsequently, in 2005, the then Martin government tabled Bill C-81 in the House of Commons to establish such a “National Security Committee of Parliamentarians.” The bill went no further than first reading in the Commons before it died on the order paper at the time of the 2006 election.

Bill C-81 has since been resuscitated 5 times in subsequent Parliaments, as a Liberal private members bill.  The most recent version -- sponsored by Wayne Easter -- is Bill C-551

Meanwhile, Senator Hugh Segal recently introduced his own private members law project -- Bill S-220 -- in the Senate.

Bills C-551 and S-220 are quite similar, although they do vary in a few, sometimes significant ways.  Both would create a statutory committee of parliamentarians -- that is, an administrative body staffed by Parliamentarians, but which is not per se a parliamentary committee.  This is a distinction with a difference.  Because the committee of parliamentarians is an administrative body and not a parliamentary committee, it would not possess parliamentary privileges (including the freedom of speech and the power to compel attendance of witnesses and papers).

In both cases, members would be appointed by the executive, without any formal full parliamentary vetting.

Both would have relatively similar and reasonably broad mandates -- to "review the legislative, regulatory, policy and administrative framework for intelligence and national security in Canada, and activities of federal departments and agencies in relation to intelligence and national security."  (Bill S-220 includes no power by a relevant minister to refer a matter to the committee.)

Both require the committee to report to the Prime Minister, who is then obliged to table the report in Parliament, with redactions of information the Prime Minister regards as "injurious to national security or national defence" and, in the case of Bill C-551, "international relations".  (Bill S-220 would require that notice be given of the existence of redactions when the report is tabled in Parliament.)

Both impose secrecy obligations on members, including rendering them persons permanently bound by secrecy under the Security of Information Act.

Both include language on limiting committee access to sensitive information.  Here, however, there is an important (indeed, fundamental) difference: Bill S-220 denies committee members Cabinet confidences, but nothing else.  Moreover, S-220 gives the committee power to compel the attendance of persons and papers. Bill C-551, in comparison, gives ministers the choice to supply information, and includes a long list of considerations ministers can take into account in declining to supply the committee with information. 

In terms of omissions, neither bill authorizes the committee to report directly to Parliament, something the Australian and New Zealand committees do.  Nor is Parliament affirmatively given the power to refer matters to the committee, something that the Australian and New Zealand parliaments may do.

Put another way, the current bills seem to draw more inspiration from the UK model, as compared to its southern hemisphere counterparts. Critics have some times questioned the UK committee model. 

It is worth, therefore, considering other systems.  For instance, the New Zealand committee (Intelligence and Security Committee) is dramatically different from what is proposed in the Canadian bills.  For one thing, the committee proceedings are parliamentary proceedings and attract parliamentary privileges. (That said, there are standards governing information disclosure to the committee.  Note though that these are much more constrained and carefully codified than the equivalents in Bill C-551.)

For another thing, the committee's membership must include a member of the House of Representatives nominated by the leader of the opposition (with agreement of the Prime Minister and after consultation with the opposition parties).  All of the nominated committee members are endorsed by Parliament, which appears able to veto nominations. 

The committee also reports to Parliament directly (and publishes its reports on the internet), subject to rules on disclosure of truly sensitive information.  The committee itself applies these confidentiality rules -- there appears to be no executive government sanitation process.

Aspects of the Australian model (Parliamentary Joint Committee on Intelligence and Security) are also worth pondering.  For one thing, its governing legislation encourages multi-party representation: members are appointed by resolution of each House of the Australian Parliament, on nomination of the Prime Minister (or government leader in the senate).  Before making these nominations, the PM and the government leader must consult with the leader of the official parties in the respective hours.  And in nominating members, the Prime Minister and government leader "must have regard to the desirability of ensuring that the composition of the Committee reflects the representation of recognised political parties in the Parliament."

For another, as noted, matters may be referred to the committee by "a resolution of either House of the Parliament".  This supplements the committee's own ability to review as it sees fit.  Also as noted, the committee reports annually to Parliament, and not through the intermediary of the Prime Minister.  Moreover, the committee reports its comments and recommendations in its reviews to each house of Parliament, as well as the responsible Minister.

This is not to suggest that the Australian model is flawless. The Australian committee has no statutory basis for extracting information from the intelligence services, and there are important qualifications and limitations on it functions (eg, functions do not include reviewing intelligence gathering and assessment priorities of the services or reach activities that do not affect an "Australian person").  It is not, therefore, a model to be emulated in its entirety.

Other models are legion, as described in a detailed 2011 EU report on the subject.  Many are more impressive than what is proposed in the two Canadian bills.

To conclude, on balance I think a Canadian committee of parliamentarians focused on national security matters would be a good thing: the more eyes on this issue, and the more institutional competence in Parliament, the better.  Bill S-220 is better than Bill C-551, especially in terms of the ability actually to obtain information from the government.  That said, both bills are less ambitious than I would like -- indeed, they replicate some of the features of the UK model that critics across the pond dislike.  (Indeed, bill C-551's sweeping language allowing the government to refuse to provide information to the committee is more aggressive than the UK equivalent and is a loophole that swallows the bill).  "Ambition" would mean looking to the southern hemisphere, not to the United Kingdom.