Privacy in Sent Text Messages

This post is in: December 21, 2017

Supreme Court of Canada settles outstanding disputes

The Supreme Court of Canada has handed down a pair of related decisions which are of great significance to the law around text messages. Both decisions settle disputes which had existed in the law until now. R v Marakah settles that a person retains a reasonable expectation of privacy in text messages after they have been sent, a point about which the British Columbia and Ontario Courts of Appeal had reached diametrically opposite conclusions. R v Jones adds a further gloss to that point, but in addition decides that a Criminal Code Part VI authorization – that is to say, an authorization to intercept private communications – is not required to obtain historical text messages: that was a point which had been left unsettled in R v TELUS Communications Co, and which had led to disagreement in courts below.

In Marakah, the accused was charged with a number of firearms trafficking offences. The police had unlawfully searched the cell phone of his accomplice, Winchester, and had obtained an incriminating text message conversation between Marakah and Winchester. The trial judge held that the accused did not have standing to object to the unlawful search of Winchester’s cell phone, admitted the text message conversation, and convicted the accused: the Ontario Court of Appeal upheld the trial judge’s conclusion. The accused renewed his argument at the Supreme Court of Canada. By a five-to-two majority, the Court concluded that the accused had a reasonable expectation of privacy in the text message conversation, and therefore did have standing to challenge the search.

Chief Justice McLachlin wrote for the majority, with Justices Abella, Karakatsanis and Gascon concurring in her decision. Although the issue was one of the accused’s standing to raise a section 8 Charter argument, the underlying question which settled that was whether Marakah had a reasonable expectation of privacy which had been violated by the search of Winchester’s cell phone. Whether a person has a reasonable expectation of privacy depends on the totality of the circumstances, which involves considering four lines of inquiry: the subject matter of the alleged search; the claimant’s interest in the subject matter; the claimant’s subjective expectation of privacy in the subject matter, and; the objective reasonableness of a subjective expectation of privacy, and so the decision looked at those factors.

A key aspect of the majority’s reasoning was in the characterization of the subject matter of the search, which really settled the analysis of the next two factors. The subject matter of the search, the majority held, should not be understood simply to be Winchester’s cell phone: rather, the subject matter of the search was better understood as the electronic conversation between Marakah and Winchester, independent of where the data happened to be stored. Once the subject matter was understood in that way, it was clear that the accused had a direct interest in that subject matter, and that he had a subjective expectation of privacy.

The other key consideration in the privacy analysis was the fourth factor, whether an expectation of privacy was objectively reasonable. That question rested on consideration of a variety of factors, the most relevant of which in these circumstances were the place where the search occurred, the private nature of the subject matter, and control over the subject matter. The Chief Justice concluded that whether the place of the search was regarded as the private electronic space where the communication between two people occurs, or as the physical device of a person or that person’s friend, there would still be some reasonable expectation of privacy in the location. Further, that subject matter – text messaging – was inherently a very private method of communication, capable of revealing intimate details of a person’s lifestyle, and preservation of a “zone of privacy” in which personal information is safe from state intrusion is the very purpose of section 8 of the Charter.

The fundamental disagreement between the majority and the dissent came down to the issue of control. The Chief Justice held for the majority that although control is a relevant factor, lack of control is not fatal to a privacy interest: the risk that a recipient could disclose a text message does not lead to the conclusion that the state can intrude upon an electronic conversation absent such disclosure. Although technological reality might mean that a person does not have exclusive control over information, that person can still reasonably expect that information to remain safe from state scrutiny.

Justice Moldaver (with Justice Côté concurring) disagreed with the majority’s conclusion, primarily based on a differing view about the significance of control. He agreed that a person need not have exclusive control or direct control over something to retain a privacy interest in it, but argued that that should be seen as different from having a total absence of control. In this case, the accused had no control whatsoever over the text messages on Winchester’s cell phone, and so did not, in the dissent’s view, have a reasonable expectation of privacy.

Chief Justice McLachlin suggests early in her decision the limited conclusion that “depending on the totality of the circumstances, text messages that have been sent and received may in some cases be protected under s. 8” (para 4), Justice Moldaver in dissent suggests that her reasoning in this case will always apply, and therefore that “the Chief Justice effectively holds that everyone has a reasonable expectation of privacy in text message conversations, even when those conversations are on another person’s phone” (para 168). Justice Moldaver sees that as a criticism of the majority decision, but it can easily be regarded as a strength of the decision rather than a weakness. Indeed, it seems reasonably clear that this was precisely the Chief Justice’s intent. What she says is that there might not be a privacy interest in all “exchange of electronic messages,” not in all text message conversations (para 5). The examples she offers later of situations which remain unsettled are “messages posted on social media, conversations occurring in crowded Internet chat rooms, or comments posted on online message boards” (para 55). A bright line rule in the particular situation of text messages conversations should be seen as a good thing.

To say “there is always a privacy interest in text messages conversations” is to provide a clear, workable and efficient rule to the police: if you want to examine a text message conversation, you need some source of lawful authority to do so. On the other hand, to say “there is sometimes a privacy interest in text message conversations” is to invite the police to search first and ask questions later. Any form of “maybe there is, maybe there isn’t” rule sends the message that on some occasions no lawful authority will be required to obtain text messages from the cell phone of another person. No-one could fault police officers for thinking that, under such a regime, this might be one of the occasions they did not need lawful authority. Even if the officers turn out later to have been wrong about whether they needed authority to search, that will only be after the person’s privacy interest has already been violated. Given that the stated purpose of section 8 of the Charter is to be “a means preventing unjustified searches before they happen, not simply of determining, after the fact, whether they ought to have occurred in the first place” (Hunter et al. v. Southam Inc., emphasis in original), that should be seen as the wrong approach.

In part R v Jones dealt with the same issue of privacy in sent text messages. In that case, the police had obtained a text message conversation which was alleged to be between Jones and a co-conspirator by means of a production order served on Telus: that production order obtained the text messages from an account associated with the co-conspirator. One of the Crown’s arguments in Jones was the same as in Marakah – the accused could have no reasonable expectation of privacy in the text messages once he had sent them to someone else – and that argument failed for the same reason as in Marakah. In Jones the text messages had been obtained from the service provider rather than the co-conspirator’s cell phone, but that did not change the analysis in any way. There was also the further wrinkle that Jones had not admitted that he had sent the text messages in question, and the Crown argued that he could not have it both ways: he either had to concede authorship or concede that he had no standing. The Supreme Court, however, held that the Crown was equally trying to have it both ways: they were claiming that Jones was the author of the texts, but also denying that that was established. The best approach, the Court concluded, was to allow Jones to argue for standing on the basis of the Crown’s allegation that he was the author.

The unique contribution of Jones arose because Jones had standing, and therefore the argument reached the stage of asking whether the search had been authorised by law. The police had used a production order in order to get the “historical” text messages: that is, ones that had already been sent and were stored in the Telus servers. Jones argued that based on the reasoning in Telus Communications obtaining text messages amounted to intercepting private communications, and therefore that only an authorization under Part VI of the Code would be sufficient.

This argument was, on the face of it, a plausible one based on R v TELUS Communications Co. In Telus Communications the police had obtained a production order which was in part like the one in Jones, in that it required Telus to hand over some text messages which had already been sent, but in addition required the production of text messages which had yet to be sent. Telus Communications only dealt with the prospective text messages, and found that a production order was not sufficient. Rather, that case held, the prospective portion of the order amounted to an interception of private communications, and so an authorization under Part VI was required. Three of the seven judges in the case had held that the prospective order was literally covered by Part VI (in a decision written by Justice Abella), while two of the seven judges had explicitly rejected that argument, and two others had expressed no opinion on that particular point.

That decision had left some ambiguity in the law. Telus Communications was as a matter of fact limited to looking at obtaining copies of text messages which had not yet been sent, and so conceptually it was easy to see it as akin to “listening in”. However, the reasoning leading to that result, at least in Justice Abella’s reasons, amounted to removing the temporal component from “intercept”. In essence, “intercept” became more or less equivalent to “obtain a copy of”. And if that were the case, then the same reasoning arguably applied to the text messages which had already been sent.

In Jones, however, Justice Côté for the majority concluded that a production order was in fact sufficient lawful authority for the search: more specifically, the police were not obliged to obtain a Part VI authorization to intercept private communications in order to obtain historical text messages. The majority concluded that although text messages are private communications, obtaining them from a service provider after they have been delivered does not constitute “intercepting” them. The Code distinguishes between “intercepting” and “disclosing” messages: the majority held that on its ordinary meaning and in the context of the statutory scheme, obtaining a copy of a text message which has been retained by a service provider does not constitute an “interception”. As a result, a Part VI authorization would only be needed where the text messages to be obtained are either not yet in existence or are still capable of delivery at the time the order is issued. Here, the police only sought text messages which had already been sent, and so the production order was sufficient authorization. As a result, the search was authorised by law and there was no section 8 violation.

Justice Abella was also on the panel in Jones and dissented on this point. In her view, the level of privacy protection for communications ought to depend on the purpose of Part VI of the Code and the character of the communication, not on the timing of the state’s request for authorization or on technological differences between service providers. The term “intercept”, she noted, has previously been given a wide meaning in this context, and there is no technological difference between obtaining a copy of a text message before or after it has been sent: the only difference is the timing of the police application. Obtaining even a historical text message either is or is substantively equivalent to an interception of private communications under Part VI, she argued, and so the production order in this case should not be seen as sufficient.

In the end, between them Jones and Telus Communications establish that there are two very different sets of rules for historical and for future text messages. Which set of rules apply depends on whether the police seek judicial authorization before the text message is sent or afterward – and in practical terms also depends on whether the target of the search uses Telus as a service provider, since other providers do not store copies of messages. That is, as Justice Rowe observes in concurring reasons, “highly anomalous” (para 84).