Robert Diab

In its decision in R v Campbell earlier this month, the Supreme Court of Canada revisited an issue it had dealt with in R v Marakah (2017), but with a twist. In that case, the Court held that a person can retain a privacy interest in a text they send, despite losing control of it in the hands of a recipient. Police had read the sender’s text when they searched the recipient’s phone incident to arrest. In this case, upon seizing the recipient’s phone, police saw new texts coming in on the lock screen and began to respond to them pretending to be the recipient.

Campbell addresses a number of issues, none of them entirely new, but all of them interesting and in need of further clarity. Did police conduct an ‘interception,’ engaging the wiretap provisions of the Criminal Code, when they hijacked a phone in this way? Were police authorized to search the phone as part of the recipient’s arrest, or were they conducting a different investigation once they took over the phone? Did the drug supplier sending the messages give police cause to carry out a warrantless search due to exigent circumstances, given his eagerness to quickly offload a quantity of heroin laced with fentanyl?

In one sense, none of this should concern anyone outside of the drug trade. But in another sense, the Court here is writing the latest chapter in a larger story about the boundaries of digital privacy, or to put it in highfalutin terms, the boundaries of the digital self.

My purpose in this post is to highlight what stands out to me about the case, having surveyed Supreme Court jurisprudence on section 8 extensively, including Marakah (2017) and Mills (2019), two cases to which Campbell forms a companion. For this reason, I won’t try to reconstruct the reasons in detail, but I will sketch the broad outlines.

How police intruded on Campbell’s privacy

Police were investigating a suspected dealer G, one of Campbell’s buyers. They obtained a warrant to search G’s residence and arrested G attempting to drive away from the house. During the arrest, he tossed two phones in the passenger seat, which police seized incident to arrest. A few minutes later, four texts appeared on the lock screen by ‘Dew’ (Campbell’s first name is Dwayne), indicating “I need 1250 for this half” and “What you gonna need that cause I don’t want to drive around with it”. Given the amount involved and their experience with the recent drug trade in Guelph, Ontario, police assumed this was likely a half ounce of heroin mixed with fentanyl, and that there was a good chance Dew would sell it elsewhere quickly, and that it was deadly. An officer began texting with Dew to arrange delivery, in an exchange that lasted over two hours involving some 35 texts. When Campbell turned up at G’s residence shortly thereafter, he was arrested and found with drugs and cash.

Campbell at trial

Campbell argued at trial that he didn’t send the first four texts, which contained the offer to supply heroin. He only sent the texts that followed, sorting out details about delivery. The decision is silent on this point, but I query whether Campbell was trying to place himself on one side of the line drawn in R v Greyeyes (SCC 1997), where the Court held that a person aiding a purchaser is not a party to trafficking. The Court also found in that case that a person who aids both the buyer and the seller is a party to trafficking. At the very least, that’s what Campbell appears to have done here.

Campbell claimed that by entering the conversation, police had interfered with his privacy and had conducted what amounted to an interception of a private communication under Part VI of the Code, requiring a warrant. The trial judge cited Marakah but held that Campbell didn’t have a reasonable privacy interest in this exchange because it “did not reveal any personal or biographical information” about Mr. Campbell and contained only “mundane comments that . . . could have been overheard on a public bus”. But even if it were private, the search was authorized under section 11(7) of the Controlled Drugs and Substances Act, which allows for a search in exigent circumstances.

Campbell on Appeal

The Ontario Court of Appeal held that Campbell did have a reasonable privacy interest in the exchange, because the texts were not “mundane,” but about a drug deal, “something one might make efforts to prevent from being overheard on a bus” – a rationale the Supreme Court would take issue with. Justice Trotter did not deal with the question of whether the exchange was an interception. He affirmed that there were exigent circumstances supporting the use of 11(7) of the CDSA. But a key facet of his ruling is his reading of Mills. A matter of dispute in the case law and commentary since Mills was decided is whether that case contains a majority holding, and if so what it holds.

Briefly, in Mills, police pretended to be a 14 year-old-girl and exchanged email and Facebook messages with the accused, resulting in charges of child luring. Justice Brown, writing for himself and Justices Abella and Gascon, found that an older man could not have a reasonable privacy interest in an online exchange with a 14-year-old girl he didn’t know. Justice Karakatsanis, for herself and Wagner CJ, held that the conversation did not involve an interference with privacy since it was not a surreptitious recording of a conversation (as in R v Duarte, 1990), but one carried out in digital rather than purely oral form (i.e., this was simply a conversation with an undercover officer, which is not a search). Justice Moldaver held that both sets of reasons were “sound in law”. Some courts and commentators had assumed this meant that the Brown opinion formed the majority.

Justice Trotter here was careful to describe the ‘majority’ opinion in Mills as comprising both Karakatsanis and Brown’s opinions (since the two together form the basis for dismissing the appeal), but Brown and Moldaver’s reasons forming the plurality opinion. Justice Trotter distinguishes what happened in Campbell from Mills by pointing to Karakatsanis’ holding that police did not interfere with a private conversation since they were in on it from the outset. Here, Trotter notes, they “insinuated” themselves into it.

Majority opinion at the SCC: was it private?

At the Supreme Court, Justice Jamal, writing for himself and Chief Justice Wagner, along with Justices Kasirer and O’Bonsawin, affirmed the appeal court’s holding but offered different reasons on two key points.

Justice Jamal held that Campbell’s exchange with police was private not because it was “something one might make efforts to prevent from being overheard on a bus” as Justice Trotter had held, but because of the nature of text messages per se. Texting is a form of communication we expect to be free from state interference altogether. And we have this expectation because they have the potential to reveal private information.

Justice Jamal’s analysis of Campbell’s privacy in the exchange largely replicates the one in Marakah, affirming the Court’s holding there that a loss of control over data or communication in a recipient’s possession does not render a sender’s privacy interest unreasonable. But the reasoning here contains fewer qualifications about texts being private. Whereas in Marakah, they can be private, the holding here comes close to asserting that texts are prima facie private, given our normative assumptions about them.

Justice Rowe agreed with Jamal’s reasons on this point without reservation, and the dissenting opinion of Justices Martin and Moreau (joined by Karaktasanis) also affirms privacy in text messages in a broad and unqualified sense (“Conversations that take place over text messaging promise participants a high degree of privacy and are capable of revealing a great deal of personal information.”) Only Justice Côté’s concurring opinion raises the point that Marakah didn’t create a ‘categorical rule’ that all texts attract a reasonable privacy interest. But given the view of eight other members of the Court, it may be that Campbell now stands for the contrary: all texts are presumptively private. And this in itself may be the most significant aspect of the decision.

The eight members of the Court (aside from Côté) also agreed that the intrusion into Campbell’s conversation was not authorized by the power to search incident to G’s arrest, since it was not ‘truly incidental’ to the reason for arresting G. The officer’s purpose in hijacking the conversation was to conduct a search of Dew for a different offence (trafficking) from those involving G. A small point, largely turning on the facts here. But one likely to be applied in future text message hijacking cases (already an emerging sub-genre as the citations in this case attest).

The Mills situation

Another tidbit that stands out to me is Justice Jamal’s discussion of Mills. A question before the Court here was whether Mills had created an exception to Marakah to the effect that where police suspect that a “relationship involves a crime,” communications are not private. Justice Jamal canvasses Mills in some detail to note that both Karakatsanis and Brown’s opinions affirmed Marakah. More to the point, Jamal was adamant that “there was no majority decision in Mills.” When Justice Moldaver asserted that both Karakatsanis and Brown’s opinions were “sound in law,” this did not result in Brown’s opinion becoming the majority holding – creating a Mills exception to Marakah. In other words, Justice Jamal was keen to reject the notion that texts are private unless the content or relationship itself is criminal in nature.

So, while Justice Jamal concludes this portion of the decision by asserting: “It is thus not necessary to decide whether Mills is properly characterized as creating an ‘exception’ to Marakah or as departing from the content‑neutral approach to s. 8 of the Charter” – what I think he’s really done is settled the debate about Mills and reduced the provocative opinions in that case to so much obiter.

Was this a wiretap?

Justice Jamal deals at some length with whether hijacking the phone constituted a wiretap, but he isn’t entirely clear about his reason for finding that it wasn’t. After canvassing the wiretap scheme and noting the Court’s holding in Telus (but not distinguishing it), he asserts: “in my view, Part VI is not engaged here because the police did not use a device employing intrusive surveillance technology.” He notes at one point that Part VI contemplates the use of “a separate ‘device or apparatus’ that effects the interception by surreptitious technological means”.

I take Justice Jamal to mean here that police did not carry out an interception of a private conversation with Campbell because while Cambell didn’t know he was talking to an officer, he knew that his speech was being recorded. I think this makes good sense. The thrust of Justice La Forest’s concern in Duarte (1990) was that the risk of being recorded by the state was a risk of a different order from your interlocutor turning out to be a tattletale. We can’t be assumed to accept the risk of being recorded whenever we speak, since the state could now be doing that at any time. But ever since we began texting, we recognized and accepted the risk that any text we might send could be shared.

Perhaps Justice Jamal assumes here that the risk of the interlocutor turning out to be a state agent is not a risk of a different order from the risk of our texts being turned over to the state by our recipient (which would also engage a privacy interest, a topic on which I happen to have written recently. Justice Jamal reiterates at the close of this segment that while police conduct may not have involved an interception under Part VI, “this was prima facie an intrusion upon Mr. Campbell’s reasonable expectation of privacy.”

To jump ahead to the dissent on this point, Justices Martin and Moreau (joined by Karakatsanis) offered a contrary view in a lengthy segment of their opinion — holding that what police did here should be construed as an interception under Part VI of the Code. Their reasons are twofold. An interception doesn’t require a separate device; Justice Jamal, they believe, misconstrues the definition in the Code. They also cite the Court’s broader definition in R v Jones (2017): an “interception relates to actions by which a third party interjects itself into the communication process in real-time through technological means.” Their second reason is implicit in their argument: while Campbell knew he was conversing in writing — and thus wasn’t being surreptitiously recorded — he didn’t know he was being recorded by the state. This latter rationale would seem to conflict with Karakatsanis’ opinion with Chief Justice Wagner in Mills, where she found the exchange with an undercover officer did not constitute an interception or a privacy intrusion because even though Mills didn’t know he was conversing with a state agent, he did know he was being recorded.

Were there in fact exigent circumstances here?

Put another way, if hijacking the phone constituted a search (since it interfered with a reasonable privacy interest), was the search reasonable? It was reasonable if it was authorized by law (R v Collins, 1987). All three levels of court affirmed that this search was authorized by law, namely section 11(7) of the CDSA.

The remainder of Justice Jamal’s majority opinion deals with the exigency question, affirming the findings below that the requirements of 11(7) were made out here. I don’t believe he adds much to the Court’s earlier decision in Paterson (2017) glossing this provision. And I don’t share the concern on the part of the dissent or by other commentators on this case that a finding of exigency here will vastly expand the scope of that power. I think the finding that there was an immediate and real public safety concern in this case largely turns on the facts. The first four texts make explicit that the Dew was in a rush to be rid of his supply. There was a good reason to believe it contained fentanyl. And the danger that someone could die that day from this supply was real.

As Justice Jamal notes, there had to be a “reasonable probability of the claimed exigency,” and in this case the trial judge’s finding on this point was sound. Police might have been able to obtain a telewarrant, the trial judge reasoned, but Dew was impatient and might have sold the drugs by then. For the majority of the Supreme Court, no error in logic here.

Justices Martin and Moreau (along with Karakatsanis) in dissent held that the circumstances did not give rise to exigency under 11(7), because the danger was only hypothetical. Dew might sell to a buyer who in turn might provide a quantity of fentanyl to a user, but whether and when that would happen was all too speculative to meet the test of a probability of an imminent risk to public safety in 11(7). And as a consequence, the majority’s holding that it did meet this test will water down the requirements of 11(7). It will allow police to carry out warrantless searches in cases involving dangerous drugs too readily – relying on the dangerousness of the drug to create a sense of urgency that isn’t there. But as Justice Jamal and Rowe make clear in their respective opinions, probable grounds to believe danger is imminent does not require a certainty or even a balance of probabilities; simply a reasonable probability. 

Again, I admire the sentiment – the sense of caution – in the Martin Moreau dissent, but I think the exigency issue in this case is largely about the particular wording of those first four texts. They were golden facts for the Crown. Not likely to be repeated. But that isn’t to say that Campbell won’t be a frequent citation for the Crown on section 8. It will also be a key case for defence.

- Homepage | Earlier Posts -