• Robert Diab (Faculty of Law

When Privacy and Terror Collide: Why New and Invasive Search Powers May Be Found Constitutional

Talk among civil libertarians and privacy advocates has been heating up over the past few months in response to a Green Paper the Liberal government released last fall, in the course of its consultation with Canadians over national security.

Buried somewhere near the end of the document was a brief discussion of powers the federal government would like to add to its counter-terror toolbox – and some of them are striking: the power to compel people to hand over passwords, to force internet service providers to install tools to intercept communications, and warrantless access to ‘basic subscriber information’ that would enable the police to connect your personal identity to a wide range of internet activity.

The prevailing view among privacy advocates has been to express concern over these powers, but to cast doubt on their likely constitutionality in light of an encouraging recent decision of the Supreme Court of Canada. In a 2014 case called R v Spencer, the Court held that Canadians have a reasonable expectation of privacy in their basic subscriber information – or roughly the metadata surrounding their internet account with Rogers or Shaw, etc. It is important to guard this information as a mean of protecting a broader interest in our anonymity online, which, in the Court’s view, has become essential to the enjoyment of privacy in the digital age. Police can therefore not obtain basic subscriber info, the Court held, without acting in exigent circumstances or pursuant to a reasonable law.

In Spencer itself, a case involving possession and distribution of child pornography, the Court took this to mean that if police lacked exigent circumstances, they needed a warrant under the Criminal Code to obtain the BSI. Many commentators have mistakenly inferred from this that Spencer means: no access to subscriber info without a warrant. But this isn’t accurate. Once again, Spencer says: no access except in exigent circumstances or where authorized under a reasonable law.

And there’s the rub.

The conversation around these new powers has drawn an undue measure of comfort from the Supreme Court’s approach to digital privacy in Spencer, without appreciating the context of that case. It was a criminal case where the only power to carry out this kind of search was a conventional warrant on the usual standard of reasonable grounds to believe an offence has been committed and evidence is likely to be found in the place to be searched. The Court, however, did not say that this standard is the only standard that might constitute a “reasonable law” in the domain of digital privacy.

Put otherwise, if the feds were to pass legislation setting out the powers it contemplates in the Green Paper – access to BSI, intercepting emails, compelling passwords – on something less than a warrant on reasonable grounds, one or more of these laws may very well be found constitutional. The reason for this is simple: it would involve an assessment of the interests to be balanced under the Charter in a different context than that of an ordinary criminal investigation.

One might dispute that there’s anything special about the national security context. (Craig Forcese, for example, points out that warrants under the CSIS Act require nothing less than reasonable grounds to believe—suggesting that in at least this one major instance involving national security, Parliament has conceded that a warrant on the usual Hunter standards is appropriate.) On this view, new powers for compelling passwords or obtaining subscriber info—even if they’re meant to thwart terrorism—should involve a warrant on nothing less than reasonable grounds, and anything more permissive should not be held to constitute a ‘reasonable law.’

But the Court may well take a different view. And we know this from its approach in many earlier cases.

To be clear, the basic framework for deciding what constitutes a “reasonable search” was sketched in the seminal case of Hunter v Southam in 1984, a case involving a conventional criminal investigation. Justice Dickson held there that a reasonable search under the Charter required a warrant on reasonable grounds, because these requirements struck the right balance between the individual’s right to be left alone and the state’s interest in prosecuting crime. He then said something that continues to shape section 8 law at its very core:

"[But…] Where the state’s interest is not simply law enforcement as, for instance, where state security is involved, or where the individual’s interest is not simply his expectation of privacy as, for instance, when the search threatens his bodily integrity, the relevant standard might well be a different one."

In other words, the Court recognized here that a rule can’t be crafted that defines for every situation what will constitute a ‘reasonable search,’ because in some cases, one or the other interest might carry more weight. This pointed the way to a host of later decisions in which courts would approve of a host of exceptions to the default requirement for a warrant on reasonable grounds. To mention just a few of the departures held to be ‘reasonable laws’ under section 8 are powers for: warrantless search using sniffer dogs of bags or lockers on reasonable suspicion, warrantless search at the border on reasonable suspicion, search incident to investigative detention on reasonable suspicion. And there are many more.

So, how might the Supreme Court weigh the interests at issue in a search power intended for terrorism investigations?

The answer can be found by pointing to the Green Paper’s central motif: a series of brief interconnected scenarios or vignettes meant to illustrate the practical utility of the powers recently added to Canada’s counter-terror law, and the need for the new ones at issue. The scenarios involve the case of a “charismatic speaker who holds weekly meetings in a local community centre” who advocates violence and influences a number of younger men and women. They’re highly emotive scenarios, triggering images of mass-casualty terror attacks, and implying unmistakably that what is stake in these powers is very different from the state interest in conventional cases like the one in Hunter v Southam.

I speculate that legislation will soon be tabled containing powers of either warrantless access to basic subscriber information or access on lesser grounds. There may also be similar powers involving passwords and the interception of email or other messages—all in the context of national security. And when the time comes to challenge these laws under section 8 of the Charter, they may well be held to be “unreasonable” as many predict. But one or more of them may not.

The Supreme Court’s record since 2001, when it comes to assessing the ‘reasonability’ of laws pertaining to national security has not been encouraging. In a number of cases where images of mass casualty terror function as a kind of unspoken premise, the Court has consistently favoured the government’s position—in substance, if not in the technical outcome of the case. These include Charter challenges to aspects of the Anti-Terrorism Act (Khawaja, App Under Section 83) and cases on deportation to torture and the security certificate regime (Suresh, Charkaoui, Harkat).

Cooler heads may prevail, but fear often shapes value judgements. And what constitutes a ‘reasonable search’ is nothing more than that.

Recommended Links

https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/ntnl-scrt-grn-ppr-2016-bckgrndr/index-en.aspx

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/14233/index.do

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2436615

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/5274/index.do

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/12768/index.do

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/2156/index.do

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/1937/index.do

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/2345/index.do

https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/13643/index.do