Summary of Concerns with CSIS Act Amendments

Kent Roach and I are preparing a series of "alternative backgrounders" on the government's troubling anti-terrorism bill.  We will be posting in reasonably short order our first, a full analysis the "advocating and promoting terrorism offences in general" offence.  It will appear on a special website created by our book publisher, and I will announce on this blog.  Those backgrounders will be developed into chapters for a book which we will produce as quickly as humanly possible.  A special plea: in this project, we want to "crowdsource" legal and practice expertise (please, not conspiracy theories and rhetoric) and will be inviting feedback -- good or bad.  My email is cforcese [at]

As we finalize that first backgrounder, I have sketched out a structure the next one, on the CSIS Act amendments.  In this blog post I summarize my current thinking:

Twinning Expansive Mandate with Expansive Powers

The CSIS mandate is extremely broad, and its reach much less precisely defined than, say, the sorts of crimes the RCMP is empowered to investigate.  A broad mandate is acceptable if the powers exercised by the Service are limited.  That is exactly the principle I have always understood animated the construction of the CSIS Act.  A democratic society should pause seriously over any proposal twinning broad mandate with expansive powers in a clandestine service.

And yet the present bill accomplishes exactly this ill.  More than that, it anticipates unlawful conduct by CSIS, done in the name of the extremely broad goal of protecting the “security of Canada”.  We need to be clear, this goes well above-and-beyond terrorism.  It is the entire orbit of national security concerns.  This is not about the October attacks, or at least not just about those attacks.

The illegality that CSIS may engage in must, in the language of the bill, be reasonable.  It stops at bodily harm, obstruction of justice and violation of sexual integrity (so no black sites for torture).  But one read of the bill’s construction is that anything short of that is permissible.  Indeed, one read of the bill is that, with judicial authorization, CSIS may not only break the regular law, but also violate the Charter.

This is extraordinary.  First, it is true that in limited circumstances the RCMP too can exceed the law.  There, however, the RCMP must be deploying their law enforcement function, investigating offences.  I do not doubt the scope of this power, but it is less expansive than the potentially highly mutable security of Canada concept in the CSIS Act, one that does allow CSIS to skate very close to investing itself in matters at the core of democracy.  

Second, when the RCMP conducts an investigation, it is typically with an eye to laying charges, and bringing a court case.  Its conduct is therefore ultimately (at least in principle) subject to disclosure.

There is no such principle for the Service.  CSIS’s conduct is rarely ever a matter of public record.  The prospect of indirect accountability through a criminal trial brought after a police investigation is almost always non existent.

For both these reasons, and probably a few others I shall develop, empowering CSIS as a lawbreaker raises serious concerns.

The Trouble with the Proposed Warrant System

I am deeply troubled by one read of the bill that seems to suggest a judge may authorize even a violation of the Charter. Yes, in the world of search and seizure, judicial warrants render police or CSIS action proper.  That is because the Charter privacy protection is not (and never has been) absolute — the violation of privacy must not be unreasonable.  Unreasonable typically means without warrant.  Put another way, the warrant requirement is conflated closely with the actual scope of the right.

Other Charter rights are dramatically different.  For instance, there is no concept of “reasonable” cruel and unusual punishment.  It is an absolute right.  To imagine that a court can pre-authorize a violation of this right is to misunderstand entirely the way our constitution works, on a fundamental level.  When a right exists, it is a space that is absolutely unavailable to government.  It is best thought of as a black hole where no government power can exist.  It is true that s.1 of the Charter does allow this space to invaded, but only when “prescribed by law” — almost always specified by statute — on top of being justified by a rigorous legal test.  A judge is in no position to invent a new exception to the Charter right.  A judge is not competent, by warrant, to prescribe by law a s.1 exception.  

Moreover, any deliberation on whether CSIS can exceed the law will be conducted in a warrant proceeding.  All these weighty legal deliberations will be done in secret, with only the judge and the government side represented.  The person affected by the illegal activity will not be there — in fact they will likely never know who visited the misfortune on them.  They cannot defend their rights.  No civil rights group will be able to weigh in.  At best (and the bill does not guarantee this, but my guess is courts will rush to do this), the public interest will be defended by a special advocate.  This is a person, able to act only by themselves, trying to balance their SA work with their busy private practice, paid a fraction of their regular wage, unable to consult often with other special advocates, sworn to secrecy, and probably fighting tooth and nail to ensure that the government is candid.  There is no equality of arms.  And more than that, the shelves have now grown with SIRC reports and Federal Court decisions complaining that CSIS has quite often failed to meet its duty of candour in closed door proceedings.

The ultimate court decision will not be public.  There will be no natural party able to appeal it.  We risk a secret jurisprudence on when CSIS can act beyond the law.  My suspicion is that the Federal Court will do its best to push out redacted versions of its case, but they will be inventing the process as they go.  So too, any appeal will depend on ad hoc arrangements, and presumably also require a very earnest special advocate with the time and inclination to press matters and involve themselves in endless novel disputes (including over whether they can actually receive remuneration for bringing an appeal).

Meanwhile, the court will not automatically know what is done under its authorization.  What is authorized and what is done by CSIS may not line up, as the Mosley J. decision suggests.  There, Mosley J only learned of the gap between authorization and conduct through accident — from a SIRC report and the report of the CSE commissioner. For a more recent discussion of the aftermath of that decision, see here.

It is these SIRC reports, therefore, that will matter most in making sure CSIS does not exceed its already troubling judicially-authorized remit to break the law.  And here, we ask about the capacity of SIRC.  

Can We Expect Much of SIRC-based Accountability in the Brave New World?

I have nothing but respect for those who work in SIRC, in very difficult circumstances.  The fact is SIRC is an underfunded, understaffed review body. Its statutory powers have not kept pace with the reality of the security service it reviews.  In 2006, the Arar Commission underscored the urgent necessity of new legislative tools allowing SIRC to coordinate with other review bodies.  But still, even now, as the security services all collaborate, the review bodies are “stovepiped” by agency.  Informal efforts to coordinate are rebuffed by the government — as I understand it, the government has even suggested that coordination would violate Canada’s criminal law on secrecy.

The recent SIRC annual report is full of indications that CSIS has not been as frank with SIRC (or the minister) as it should be.

The government eliminated the inspector general of CSIS (supposedly for financial reasons).  We lost something like 8 accountability people there.  As a sop, the government gave SIRC something like one more staff person, from what I understand. (I have more on this here)

SIRC needs more money, more people, a robust and credible process of appointing SIRC members (in its members, it has been both understrength and indifferently staffed for some time), and a renewed mandate that is recrafted to reflect reality per the Arar Commission.  I welcome the appointment of Dean Holloway.  But the body remains below full strength, and even at full strength, members are part time.  One more part-time appointment alone will not solve any problems.

I have a piece on my concerns with SIRC’s prospects here.

I also have some statistical analyses of SIRC here.

Accountability More Generally

Let me say more on accountability: More generally, when it comes to review, we have huge swathes of security agencies with no review whatsoever.  CBSA is an extremely powerful body that effectively does both intelligence and law enforcement — but here, there is no review whatsoever (other than the occasional coroner's inquiry when someone dies in CBSA custody).

And there is no pinnacle to the review system.  No accountability agency sees the “big picture”.   They operate in their own corners, siloed from one another.

The government's current line about more accountability being about red tape is a misrepresentation.  In the  past, CSIS used to say that the SIRC has made it a better service.  If that is so, a SIRC that can actually accomplish the objectives it was originally designed for is not “red tape” — it is an investment.  

“Pinnacle” or “all of government” accountability in national security is precisely the most compelling reason for a committee of parliamentarians.  To point to a vacuum and say that filling it would be redundant, as the government seems to be doing, is an argument persuasive only to those inattentive to reality.

And to say that there is something uniquely inept about Canada — as opposed to basically every other western democracy — and that we cannot manage effectively both bureaucratic and parliamentary review of security does us little credit.

In this bill, the government proposes radically restructuring CSIS and turning it in a “kinetic” service — one competent to act beyond the law.  This is rupture from the entire philosophy that animated the CSIS Act when it was introduced 30 years ago.  I do not believe it is warranted.  I especially do not believe it to a reasonable course without a redoubled investment in our tattered accountability system.  Anyone who has worked on accountability in the security sector knows that the core maxim is “trust but verify”.  The Prime Minister, in condemning those raising doubts about accountability, asks us simply to trust.