Secrets and Strategic Leaking: Situating Canada among its close allies

Spring has sprung, and so has the United States intelligence community. Specifically, the last several months have witnessed an astonishing pattern of leaks from the US government.  Some of the leaks come from the White House -- in one case apparently as part of run-on bombast from the man at the top. But in the wake of the Manchester bombing last week, some leaks seem to be coming out of the intelligence community itself. American leaks have fuelled angst among allies, most recently the British.

The New York Times ran two interesting analyses of US leaking of information collected by British authorities in the Manchester bombing investigating (here and here). They note how the culture of leaking is embedded in the United States in a manner without parallel in the United Kingdom, not least because of the broad reach of the First Amendment. I suspect also that the competing power centres in the American republican system (Congress/Executive) also facilitate an environment receptive to leaking. And the sheer size of the US intelligence community, and its vast penumbra of contractors, has contributed to leaky information control (the Snowden matter constitutes the obvious poster child for this).

But where does (Westminster) Canada lie on a spectrum of leakiness, with the United Kingdom on one pole and the United States on the other? My instinct is to say we are closer to the United Kingdom than the United States, with caveats.

The Law

First, there are many serious penalties for leaking classified information. On top of workplace discipline and loss of a security clearance, there is a prospect of going to jail. Security and intelligence community employees (and a number of others) are "persons permanently bound by secrecy" under the Security of Information Act (SOIA). It is a serious crime for them to reveal special operational information.  It could also be a breach of trust under the Criminal Code for any official to leak any classified information. And s.18 of the CSIS Act includes its own offence for leaks of Service information relating to source identity, applicable to those acting under the CSIS Act (a category that would include not just CSIS officers and employees, but also those of SIRC and ministerial officials.)

US and UK law also include penalties for leaky intelligence employees.

But we are more like the United States in one key respect: our Security of Information Act is probably even more unworkable than the US Espionage Act (and way more unworkable than the UK Official Secrets Act) when it comes to leaks by officials, other than those permanently bound by secrecy, to entities, other than foreign entities or terrorist groups. So, for instance, a leak by an official (other than a person permanently bound by secrecy) to the Canadian media would be very difficult to prosecute under the Security of Information Act (although Criminal Code breach of trust would still apply).  The reason for this is that the "anti-leakage" provision of SOIA (the infamous section 4) has not been updated since 1939 and is unconstitutional gibberish -- and so concluded the Ontario Superior Court of Justice in 2006.

We are also more like the United States in our constitutional arrangement. Pointing to Charter s.2, the Federal Court has recognized a constitutionalized right in Canada of whistleblowing. But this right is limited by s.1 considerations (See discussion in Forcese & Freeman, Laws of Government, 2011 at 220). And basically, that means that whistleblowing is limited to dire necessity, and must be done first through an internal mechanism and not initially to the media and public.

This approach is codified even for persons permanently bound by secrecy as a defence to a criminal charge under the SOIA, where they might otherwise be convicted for leaking special operational information. That defence specifies that whistleblowing should go through SIRC or the CSE commissioner, although the provision (s.15) leaves a lot of unanswered questions about how that mechanism might work.


On balance, we seem to be more like the UK in our practice. It is true that Canada has management issues with classified information. For instance, the CBC reported in November 2016 that "[t]here have been more than 10,000 incidents of classified or secure documents being improperly left or stored since Prime Minister Justin Trudeau's government came to office." This included 659 cases in CSIS itself. But all countries struggle with information management -- that is different from intentional leaking.

And CSIS has complained regularly about compelled disclosure in security certificate and Canada Evidence Act proceedings. But again, disclosure by court order is different from leaking.

In relation to intentional, unauthorized disclosure (aka leaking) of intelligence by the intelligence community or related entities (such as review bodies or special advocates), I cannot think of many (publicized) examples from the recent past.

The two instances that do come to mind were concerning, not because they revealed sensitive secrets but because they constitute examples of someone in government selectively disclosing intelligence information (or innuendo at least) to cast the security services in a favourable light.

The first was the leak concerning Maher Arar, during the Arar inquiry. This involved both selective leaks of information/analyses along with slanderous editorializing about Arar's character. And it came just as the RCMP and CSIS (among others) were being rightly castigated for their failings in the treatment of Arar. The Arar commission was so concerned about this behaviour, it added an entire chapter on the leaks to its report. And it said this about leaks done in 2003:

Unlike many other actions of Canadian officials that I describe in this report, leaking information is a deliberate act. Moreover, some of the leaks relating to Mr. Arar were purposefully misleading in a way that was intended to do him harm. It is disturbing that there are officials in the Canadian public service who see fit to breach the public trust for their own purposes in this way. It is disappointing that, to date, no one has been held accountable.


The second major example were similar leaks relating to Messieurs Charkaoui and Abdelrazik, again as the government was losing in front of a judge while trying to defend its efforts to deport and exile, respectively, these men.

It is unclear who did all this leaking -- a lot of the information was in circulation in government departments outside of the RCMP and CSIS. Both of these leak cases prompted RCMP investigations -- none have resulted in any further action being taken. In the Arar matter, the RCMP's hamfisted investigation involved a raid on Ottawa Citizen journalist Juliet O'Neill's home and office.  That in turn prompted the Citizen to challenge the constitutionality of s.4 of the Security of Information Act -- successfully.

In the second matter, an investigation has apparently been ongoing, but here again the RCMP appears to have made serious strategic errors by targeting journalists in an effort to find their sources. And on top of that, they did so in a manner that did not comply with ministerial directions on national security investigations that implicate sensitive sectors (such as the media).

I discuss the fall-out from this conduct here.


So in sum: We are like the United Kingdom in our strict rules and a general pattern of few intelligence leaks from our broader intelligence and security community.  We are like the United States in that leaking raises constitutional free speech issues (and also constitutional safeguards on overbroad anti-leakage provisions, like s.4 of SOIA).  And we have an unpleasant tradition -- however modest -- of government leaks to the media designed to cast the security services in the most favourable light when those services are embroiled in controversy.