Supreme Court of Canada

Appeal Heard: May 17, 2022 Judgment Rendered: November 24, 2022 Docket: 39664

Between: Corey Daniel Ramelson — Appellant and His Majesty The King — Respondent — and — Director of Public Prosecutions, Criminal Lawyers' Association of Ontario, British Columbia Civil Liberties Association and Canadian Civil Liberties Association — Interveners

Coram: Wagner C.J. and Moldaver, Karakatsanis, Côté, Brown, Rowe, Martin, Kasirer and Jamal JJ.

Reasons for Judgment: (paras. 1 to 101) Karakatsanis J. (Wagner C.J. and Moldaver, Côté, Brown, Rowe, Martin, Kasirer and Jamal JJ. concurring)

Indexed as: R. v. Ramelson

2022 SCC 44

File No.: 39664.

2022: May 17; 2022: November 24.

Present: Wagner C.J. and Moldaver, Karakatsanis, Côté, Brown, Rowe, Martin, Kasirer and Jamal JJ.

on appeal from the court of appeal for ontario

Headnote

Criminal law — Abuse of process — Entrapment — Bona fide inquiry — Virtual space — Internet — Accused responding to ad posted by police in escort section of online classified advertising website — Undercover officer posing as escort disclosing to accused in ensuing text message chat that she was underage — Accused arrested when attending at hotel room to meet officer and charged with child luring‑related offences — Accused convicted but seeking stay of proceedings on basis of entrapment — Whether police had reasonable suspicion that criminal activity was occurring in space defined with sufficient precision — Whether police entitled to offer opportunity to commit child luring offences — Application of entrapment framework to online police investigations.

"Project Raphael", an online investigation conducted by the York Regional Police, led to the arrests of 104 men between 2014 and 2017 for child luring and related offences. Ads posted by the police on the escort subdirectory of Backpage.com spurred text‑message conversations, where an undercover officer, after agreeing to provide sexual services, revealed themselves to be underage. All those who took up the invitation to meet the undercover officer in a designated hotel room were arrested. In 2017, R responded by text message to such an ad posted by an undercover office posing as "Michelle", who revealed she and her "young friend" were 14, after having agreed to a transaction with R. When R arrived at the hotel room, he was arrested and charged with three offences: (1) child luring under 16 (s. 172.1(1)(b) of the Criminal Code); (2) communicating to obtain sexual services from a minor (s. 286.1(2)); and (3) arrangement to commit sexual offences against a person under 16 (s. 172.2(1)(b)).

R was convicted of all three offences at trial, but applied to stay the proceedings based on entrapment. Entrapment occurs where police provide a person with an opportunity to commit an offence without reasonable suspicion that this person is already engaged in criminal activity; or without reasonable suspicion that crime is occurring in a sufficiently precise space (i.e. not acting pursuant to a bona fide inquiry). The application judge initially dismissed the application. However, after the Court released its decision in R. v. Ahmad, 2020 SCC 11, where it considered how the principles of bona fide inquiries apply in virtual spaces, the judge asked for further submissions. He then concluded R had been entrapped because the virtual space was too broad to support police's reasonable suspicion and the police lacked reasonable suspicion over R personally. The Court of Appeal allowed the Crown's appeal and set aside the stay of proceedings, concluding that the application judge erred in finding Project Raphael was not a bona fide inquiry. The application judge failed to consider other relevant Ahmad factors in assessing whether the virtual space was sufficiently precise.

Held: The appeal should be dismissed.

R was not entrapped. First, Project Raphael was a bona fide inquiry. The police had reasonable suspicion over a space defined with sufficient precision. Here, the space was the particular type of ads within the York Region escort subdirectory of Backpage that emphasized the sex worker's extreme youth. Second, the offences offered by the police were rationally connected and proportionate to the offence they reasonably suspected was occurring in that space.

Entrapment is not a defence, but a form of abuse of process whose only remedy is a stay of proceedings. The entrapment doctrine strives to balance competing imperatives: on the one hand, the rule of law and the need to protect privacy interests and personal freedom from state overreach, and on the other hand, the state's legitimate interest in investigating and prosecuting crime. When the police lack reasonable suspicion that an individual is already engaged in criminal activity, the entrapment doctrine forbids them from offering opportunities to commit offences unless they do so in the course of a "bona fide inquiry": that is, where they (1) reasonably suspect that crime is occurring in a sufficiently precise space; and (2) have a genuine purpose of investigating and repressing crime. Satisfying those criteria entitles the police to present any person associated with the space with the opportunity to commit the particular offence — even without individualized suspicion of the person investigated. That test applies to investigations in physical and virtual spaces alike. However, online spaces differ from physical spaces in significant ways. The Internet, being a space that is informational rather than geographical, sheds many of the physical world's limitations in terms of scale and functions. People often behave differently online than they would in the physical world. Virtual spaces also raise unique rights concerns when they are the target of state surveillance or investigation, since they may reveal vast amounts of highly personal information. The greatest consequence of these differences for bona fide inquiries is that the boundaries of an online "space" only tell part of the story in determining whether the space is sufficiently precise. Given the potential of online investigations to impact many more individuals than an equivalent investigation in a physical space, courts assessing whether an online police investigation was a bona fide inquiry must pay close attention to the virtual space's functions and interactivity. How the police act on the Internet may matter as much or more as where they act.

Reasonable suspicion of particular criminal activity, in bona fide inquiries, must be grounded in a particular space. When that space is virtual, it is critical that the police carefully delineate and precisely define the space where they reasonably suspect crime is occurring, to ensure they have narrowed their scope so that the purview of their inquiry is no broader than the evidence allows. In Ahmad, the Court listed six factors that may illuminate the assessment of whether the police investigation was properly tailored: (1) the seriousness of the crime in question; (2) the time of day and the number of activities and persons who might be affected; (3) whether racial profiling, stereotyping or reliance on vulnerabilities played a part in the selection of the location; (4) the level of privacy expected in the area or space; (5) the importance of the virtual space to freedom of expression; and (6) the availability of other, less intrusive investigative techniques. These factors are contextual and non‑exhaustive, and no one factor should be allowed to overwhelm the analysis. The space, the crimes and the nature of the investigation all influence the acceptable scope of the police's inquiry. The entire context, in short, determines whether the space of an investigation was sufficiently precise.

Scrutiny into whether a space is sufficiently precise is crucial when police investigate broader spaces like a website. First, in permeable and interactive spaces, the "space" of an inquiry will not necessarily be intuitive. The possibility of creating subspaces, such as postings within a broader website, suggests that the ways subspaces are embedded in broader online spaces may be critical for understanding how the space of an inquiry was tethered to reasonable suspicion. Second, whether an online space was sufficiently precise may turn as much on the space's functions and interactivity as it does on its parameters. A space's functions may require that police further tailor the location of an online inquiry. They may require the police to focus on more carefully delineated spaces and target their opportunities to particular subspaces or to particular ways in which users engage with the virtual space. They may also call for attention to how the space facilitates or inhibits data collection. The Ahmad factors may assist in this determination. Entire websites will rarely be sufficiently particularized, as multi‑functional virtual spaces will usually be too broad to support reasonable suspicion. But in some virtual spaces, the criminality may be so pervasive that it supports a reasonable suspicion over the entire area. In sum, the internet's unique features are inescapable in assessing whether the location is sufficiently precise to ground reasonable suspicion.

In the instant case, the application judge erred in finding the space was insufficiently precise to ground reasonable suspicion. First, the police had reasonable suspicion that the s. 286.1(2) offence was occurring within ads posted in the escort subdirectory of the York Region Backpage based on testimony of the undercover officer which was grounded in his direct and indirect experiences in law enforcement. Second, this reasonable suspicion was related to a space that was sufficiently precisely defined. The investigation did not extend to an entire website as the space was the particular type of ads within the York Region escort subdirectory of Backpage that emphasized the sex worker's extreme youth. The relationship between the user‑created ads (the location where the police's suspicion first arose) and the police‑created ads (where the police later offered their opportunities) is integral to defining the space. It explains how the police‑created ads could be premised upon and tethered to reasonable suspicion. The space's functions and interactivity permitted the police to design Project Raphael in a way that narrowed the investigation's scope. Although the investigation impacted many individuals, in context, the purview of the police inquiry went no broader than the evidence allowed.

Finally, bona fide inquiries do not restrict police to offering opportunities to commit only the same offences that the police suspect are occurring in the space they are investigating. Following R. v. Mack, [1988] 2 S.C.R. 903, the crime that police offer must be rationally connected and proportionate to the offence they suspect is occurring. In the present case, while the police lacked reasonable suspicion over the child luring offences under ss. 172.1(1)(b) and 172.2(1)(b), these offences were rationally connected and proportionate to the s. 286.1(2) offence. The three offences capture similar conduct, have shared elements, and the luring offences are not a disproportionality more serious crime than the s. 286.1(2) offence as their sentences remain comparable.

Cases Cited

Applied: R. v. Ahmad, 2020 SCC 11; considered: R. v. Barnes, [1991] 1 S.C.R. 449; referred to: R. v. Jaffer, 2022 SCC 45, [2022] 3 S.C.R. 496; R. v. Haniffa, 2022 SCC 46, [2022] 3 S.C.R. 504; R. v. Dare, 2022 SCC 47, [2022] 3 S.C.R. 510; R. v. Mack, [1988] 2 S.C.R. 903; R. v. Conway, [1989] 1 S.C.R. 1659; R. v. Babos, 2014 SCC 16, [2014] 1 S.C.R. 309; R. v. Jewitt, [1985] 2 S.C.R. 128; R. v. Hamilton, 2005 SCC 47, [2005] 2 S.C.R. 432; R. v. Spencer, 2014 SCC 43, [2014] 2 S.C.R. 212; R. v. Chehil, 2013 SCC 49, [2013] 3 S.C.R. 220; R. v. Stairs, 2022 SCC 11, [2022] 1 S.C.R. 169; R. v. Stack, 2022 ONCA 413; R. v. Nelson, 2021 BCCA 192; R. v. Ghotra, 2020 ONCA 373, 455 D.L.R. (4th) 586, aff'd 2021 SCC 12, [2021] 1 R.C.S. 521; R. v. Levigne, 2010 SCC 25, [2010] 2 S.C.R. 3; R. v. Legare, 2009 SCC 56, [2009] 3 S.C.R. 551; R. v. Morrison, 2019 SCC 15, [2019] 2 S.C.R. 3; Kirzner v. The Queen, [1978] 2 S.C.R. 487; Amato v. The Queen, [1982] 2 S.C.R. 418; R. v. Friesen, 2020 SCC 9.

Statutes and Regulations Cited

Criminal Code, R.S.C. 1985, c. C‑46, ss. 152, 172.1, 172.2, 286.1(2).

Authors Cited

Roach, Kent. "Entrapment and Equality in Terrorism Prosecutions: A Comparative Examination of North American and European Approaches" (2011), 80 Miss. L.J. 1455.

Tanovich, David M. "Rethinking the Bona Fides of Entrapment" (2011), 43 U.B.C. L. Rev. 417.

APPEAL from a judgment of the Ontario Court of Appeal (Juriansz, Tulloch and Paciocco JJ.A.), 2021 ONCA 328, 155 O.R. (3d) 481, 406 C.C.C. (3d) 1, 72 C.R. (7th) 301, [2021] O.J. No. 2626 (QL), 2021 CarswellOnt 7056 (WL), setting aside the stay of proceedings entered by de Sa J., 2020 ONSC 5030, 67 C.R. (7th) 96, [2020] O.J. No. 4396 (QL), 2020 CarswellOnt 14800 (WL). Appeal dismissed.

Richard Litkowski and Myles Anevich, for the appellant.

Lisa Fineberg and Katie Doherty, for the respondent.

David Quayat and Chris Greenwood, for the intervener the Director of Public Prosecutions.

Michael Lacy and Bryan Badali, for the intervener the Criminal Lawyers' Association of Ontario.

Gerald Chan and Spencer Bass, for the intervener the British Columbia Civil Liberties Association.

Danielle Glatt and Catherine Fan, for the intervener the Canadian Civil Liberties Association.

Reasons for Judgment

The judgment of the Court was delivered by

Karakatsanis J. —

I. Overview

[ 1 ] Some of the most pernicious crimes are the hardest to investigate. To draw those crimes into the open, the police, acting undercover, sometimes create occasions for people to commit the very crimes they seek to prevent. Done properly, such techniques may cast new light on covert offending, unveiling harms that would otherwise go unpunished. But taken too far, they may tempt the vulnerable or the morally wavering into criminality, and virtue-test many others, threatening privacy and the public's confidence in the justice system. They demand caution.

[ 2 ] The stakes are highest on the Internet. While the medium has made activities more efficient, widespread, and harder to track, it has also allowed state surveillance to become, potentially, ever more expansive. The dilemmas this creates for balancing law enforcement with civil liberties, the rule of law, and the repute of the justice system are ongoing. This appeal, and its companion appeals, raise one of them.

[ 3 ] Between 2014 and 2017, "Project Raphael", an online investigation of the York Regional Police (YRP), led to the arrests of 104 men for child luring and related offences. Ads posted by the police on the escort subdirectory of Backpage.com spurred text-message conversations, where an undercover officer, after agreeing to provide sexual services, revealed themselves to be a juvenile. All those who took up the invitation to visit the designated hotel room were arrested. Among them was the appellant in this case, Mr. Ramelson, as well as the three appellants in the related appeals (Mr. Jaffer (R. v. Jaffer, 2022 SCC 45, [2022] 3 S.C.R. 496), Mr. Haniffa (R. v. Haniffa, 2022 SCC 46, [2022] 3 S.C.R. 504), and Mr. Dare (R. v. Dare, 2022 SCC 47, [2022] 3 S.C.R. 510)). They argue they were entrapped.

[ 4 ] When the police lack reasonable suspicion that the individual is already engaged in criminal activity, the entrapment doctrine forbids them from offering opportunities to commit offences unless they do so in the course of a "bona fide inquiry": that is, where they (1) reasonably suspect that crime is occurring in a sufficiently precise space; and (2) have a genuine purpose of investigating and repressing crime (R. v. Ahmad, 2020 SCC 11, at para. 20). That test applies to investigations in physical and virtual spaces alike. But as this Court noted in Ahmad, "state surveillance over virtual spaces is of an entirely different qualitative order than surveillance over a public space" (para. 37). There, the Court considered those differences in the context of surveillance that transpired in the investigative "space" of a phone number. This appeal, and the three related appeals, require us to do the same in the context of the Internet.

[ 5 ] At its core, the entrapment doctrine recognizes that sometimes "the ends do not justify the means" (R. v. Mack, [1988] 2 S.C.R. 903, at p. 938). Given the Internet's potential reach, there is a strong public interest in ensuring that online police investigations do not unduly intrude on public life. In assessing whether an online space is sufficiently precise to ground the police's reasonable suspicion, then, the Internet's unique features must be considered. Being informational rather than geographical, online spaces flout the limitations of physical spaces; they may lead people to behave differently than they do in person; and their use can raise distinct rights concerns, notably over privacy. Unlike physical spaces, an online space's parameters may say little about whether the space of an investigation was sufficiently precise. Instead, the space must be viewed with particular attention to its functions and interactivity to ensure that the space has been "carefully delineate[d] and tightly circumscribe[d]" (Ahmad, at para. 39). The factors discussed by this Court in Ahmad — in particular, the number of activities and people affected, the interests of privacy and free expression, and the availability of less intrusive investigative techniques — may assist in that assessment. They may be key to ensuring that the purview of an online police investigation was no "broader than the evidence allow[ed]" (para. 41).

[ 6 ] Applied here, I agree with the Court of Appeal for Ontario that the application judge erred by failing to consider factors beyond the number of people affected by the police investigation. On the correct analysis, the police had reasonable suspicion over a sufficiently precise space and the offences the police offered were rationally connected and proportionate to the offence they reasonably suspected was occurring. Mr. Ramelson was therefore not entrapped. I would dismiss the appeal.

II. Facts

A. Project Raphael

[ 7 ] Like much else, the market for juvenile sex work migrated to the Internet over the past decade. Recognizing a need to adapt their techniques, the YRP became a Canadian leader in efforts to address the issue proactively — by searching for crimes, rather than waiting for them to be reported. Whether those efforts went too far, and crossed from legitimate investigation into entrapment, is the question in this case.

[ 8 ] Inspector Thai Truong joined the YRP as an officer in 2002, and joined its drugs and vice enforcement unit soon thereafter. From 2008 onward, he investigated the commercial sexual exploitation of young girls and women. That assignment coincided with a shift in the YRP's approach. With little experience investigating juvenile sex work offences before 2008 — despite the many anecdotal signs that they were occurring — the police realized, in Insp. Truong's words, "that unless you really look for it, you're not going to find it" (A.R., vol. II, at pp. 24-25). They began educating themselves on the problem, surveying its prevalence, and, eventually, designing investigations that were adapted to its clandestine nature.

[ 9 ] Juvenile sex work found many homes on the Internet, and lacking the means to pursue all of them, the YRP's focus eventually settled on the escort subdirectory of Backpage. As a forum dedicated to the sexual services market, with hundreds of ads per night in the Greater Toronto Area alone, the volume of illegal activity on the platform was immense. And much of it was apparently underage, a reality reinforced for Insp. Truong at every turn, from professional conferences; contacts with community groups and non-governmental organizations; and the dozens of investigations into juvenile sex work — including interviews with hundreds of sex workers — that he assisted in over his career.

[ 10 ] Proactive investigations into sex work can target either the "seller side" or the "buyer side" of the market. In December 2013, the YRP launched a seller-side vice probe that, in the 2 weeks it was live, identified 31 sex workers in the York Region subpage on Backpage who appeared to be juveniles, 9 of whom were in fact underage. It also found that those who came in contact with the investigation first started selling sex, on average, at 14.8 years old.

[ 11 ] While insightful, such investigations were, to Insp. Truong's mind, only a partial success. Because selling sex is not itself illegal, the police depended upon sex worker cooperation. But whether fearing their pimps, anxious for their livelihoods, lacking a home to return to, or refusing to see themselves as victims, sex workers were often reluctant to assist. Locating the juveniles had done little to abate the juvenile sex work market. So, changing course, Insp. Truong devised a buyer-side investigation called "Project Raphael", which the YRP launched in 2014.

[ 12 ] With few models to go on, Insp. Truong was inspired by an investigation in British Columbia, where undercover police, posing as juveniles, placed ads on Craigslist offering sex workers aged 18, with descriptors suggesting youth. Project Raphael placed similar ads on Backpage, listing the age as 18 (the minimum the website would permit) and using words like "tight", "young", "new" or "fresh" in the ad's text, emulating common Backpage advertisements for the youngest sex workers (2019 ONSC 6894 (first ruling on entrapment), at para. 11; A.R., vol. II, at p. 135). When potential clients responded, the police, imitating an adolescent's idiom, arranged a sexual transaction. When the client agreed, the police revealed the sex worker was underage. When the client continued to engage, the police invited them to a hotel room. And when the client opened the door, the police arrested them.

[ 13 ] The police were aware that this design would capture more than just the most determined sexual predators. As Insp. Truong explained, those who clicked on the police's ads could be divided into three groups: (1) those "who were looking strictly for adult females"; (2) those who "were not specifically looking but were open to engaging in sex with underage females when the opportunity presented itself"; and (3) those "who contacted the ad specifically looking to have sex with an underage female" (first ruling on entrapment, at para. 14). The project's focus was on the second and third groups. But the only way to tell which category a buyer fell into, on Insp. Truong's account, was to communicate with them. And the project never attempted to track or categorize the responses.

[ 14 ] The investigation occurred in four phases, whose form evolved over time. The first phase, in 2014, listed an email address with the number "16" as a further clue to the poster's age, but the flood of responses so deluged the officers that they restricted communications to text thereafter. For similar reasons, the police eventually lowered the disclosed age from 16 to 14, to concentrate resources on the most serious offences. Intentionally or not, this had the effect of exposing those caught to more serious charges for luring minors under 16.

[ 15 ] Although never recorded, the number of responses was "overwhelming". And the number of arrests was significant. In 2014-15, posing most often as a 16-year-old, the police made a total of 32 arrests in 8 days online. In 2016, with the age lowered to 15, the police made 53 arrests in 8 days. And in 2017, with the age further lowered to 14, the police made 19 arrests in 4 days. In total, Project Raphael led to the arrest of 104 people, all in only 20 days of operation.

B. Corey Daniel Ramelson

[ 16 ] Mr. Ramelson was among those arrested in 2017. On March 27, he messaged "Michelle", aged 18, who was described as a "Tight Brand NEW girl . . . who is sexy and YOUNG with a tight body", with a "YOUNG FRIEND if your [sic] interested too" (A.R., vol. I, at p. 130). The ad featured three faceless photographs of an undercover officer in her 30s, wearing a t-shirt from a local high school. After 27 minutes of somewhat sporadic conversation, and having agreed to a transaction, the undercover officer (UC) revealed their "true" ages:

[16:28 – UC]: Just so you know we under 18. Some guys freak out and I don't want problems. We are small and it's obvious.

[16:29 – Ramelson]: I'm cool with it. I'll be gentle as long as you're sexy and willing

[16:31 – UC]: We are both willing. We're 14 but will both be turning 15 this year. That cool? We are buddies and very flexable [sic]??

[16:32 – Ramelson]: Should be lots of fun

[16:32 – Ramelson]: Are those read [sic] photos from the ads. Those girls look a bit older

[16:36 – UC]: They are both us.

[16:37 – Ramelson]: Ok. I'm going to leave now

(A.R., vol. I, at pp. 133-34; see also first ruling on entrapment, at para. 20.)

[ 17 ] Arriving at the hotel room two hours later, Mr. Ramelson was arrested. He was charged with three offences:

• Telecommunicating with a person he believed was under the age of 16 years for the purpose of facilitating the commission of an offence under s. 152 of the Criminal Code, R.S.C. 1985, c. C-46 (invitation to sexual touching) contrary to s. 172.1(1)(b) (child luring under 16);
• Communicating for the purpose of obtaining for consideration the sexual services of a person under the age of 18 years, contrary to s. 286.1(2) (communicating to obtain sexual services from a minor); and
• Telecommunicating to make an arrangement with a person to commit an offence under s. 152 (invitation to sexual touching) contrary to s. 172.2(1)(b) (arrangement to commit sexual offences against a person under 16).

[ 18 ] He was tried and convicted of all three offences in 2019 (2019 ONSC 4061). He then applied to stay proceedings on the basis of entrapment.

III. Judicial Decisions

A. Rulings on Entrapment Application, Ontario Superior Court of Justice (de Sa J.)

[ 19 ] Justice de Sa initially dismissed Mr. Ramelson's entrapment application in November 2019. While the police lacked reasonable suspicion over Mr. Ramelson specifically, they "had a reasonable basis to believe that individuals" — whether actively seeking juveniles or not — were "routinely involved in the purchase of sexual services from juvenile prostitutes on Backpage.com" (first ruling on entrapment, at paras. 51-52). Project Raphael was thus a bona fide inquiry. He rejected Mr. Ramelson's further argument that he had been induced.

[ 20 ] This Court then released its decision in Ahmad, where it considered how the principles of bona fide inquiries apply in virtual spaces. The judge invited the parties to make further submissions. And in separate reasons, issued in October 2020, he revised his conclusion: Mr. Ramelson had been entrapped.

[ 21 ] Backpage, he now concluded, was too broad a space to support reasonable suspicion. The website was not "dedicated to underage prostitution" — indeed, the "overwhelming majority" of the activity in even its escort subsection did not involve juveniles (2020 ONSC 5030, 67 C.R. (7th) 96 (second ruling on entrapment), at para. 24). Given the age listed, and the photos in the ad, "there would be no reasonable basis to infer the caller contacting the ad would be looking for an underage girl" (para. 27 (emphasis in original)). Nor did any such inference arise from the texts, and the "bait and switch" of announcing the age late in the conversation raised "clear entrapment concerns" (para. 29). Since the language in the ads was "known by police to draw in a much broader pool of individuals than simply individuals looking for 'underage' girls" (para. 31), the officer should have done more to confirm that Mr. Ramelson was himself looking for one before inviting him to commit crimes. Lacking reasonable suspicion over the space, or over Mr. Ramelson personally, the police's offer amounted to entrapment.

B. Court of Appeal for Ontario, [2021 ONCA 328](https://www.minicounsel.ca/oca/2021/328), 155 O.R. (3d) 481 (Juriansz, Tulloch and Paciocco JJ.A.)

[ 22 ] The Crown's appeal from the application judge's decision to stay Mr. Ramelson's conviction was heard together with three defence appeals from conviction (involving the appellants in the companion appeals before this Court in Jaffer, Haniffa and Dare). Concluding that the application judge in Mr. Ramelson's case had erred in finding Project Raphael was not a bona fide inquiry, the Court of Appeal allowed the Crown's appeal and set aside the application judge's order. In three separate decisions, the court also dismissed the related defence appeals.

[ 23 ] In Mr. Ramelson's matter, Juriansz J.A. explained that it was open to the application judge to find that the police had reasonable suspicion that the criminal activity under investigation was taking place on the escort section of Backpage (the offences under ss. 286.1(2) and 172.1(1)(b)). First, while the YRP lacked detailed statistics, Insp. Truong, who had extensive experience with the offences, gave sufficient evidence to show police had a reasonable suspicion that customers were going onto Backpage to obtain sexual services from persons they knew or believed to be under 18 contrary to s. 286.1(2).

[ 24 ] And while the same could not be said for the child luring offence in s. 172.1(1)(b), which concerned a person believed to be 16 rather than 18, that offence was nonetheless rationally connected and proportionate to the s. 286.1(2) offence. Given their shared elements, including the use of telecommunication, the police, in posing as a juvenile, would inevitably offer an opportunity to commit a child luring offence under s. 172.1(1) when offering an opportunity to commit the s. 286.1(2) offence. Since the police had reason to believe that juvenile sex workers, on average, first started selling sex under the age of 15, the offences of child luring under 16 (s. 172.1(1)(b)) and child luring under 18 (s. 172.1(1)(a)) were both rationally connected to the s. 286.1(2) offence. And child luring under 16 is not "much more serious", nor "totally unrelated" to the s. 286.1(2) offence (para. 89).

[ 25 ] However, the application judge erred in assessing whether the virtual space was sufficiently precisely defined. In particular, he failed to consider other relevant factors beyond the number of people affected by the investigation: including, the "number and nature of activities affected, the nature and level of the privacy interest affected, and the importance of the virtual space to freedom of expression" (para. 97). Reviewing the relevant Ahmad factors anew, the court concluded that Project Raphael was a bona fide inquiry and Mr. Ramelson had not been entrapped.

IV. Issues

[ 26 ] The case raises two broad issues:

• How does the bona fide inquiry prong of the entrapment doctrine apply in the context of online police investigations?

• Did the application judge err in concluding that Mr. Ramelson was entrapped?

  (i) Did the police have reasonable suspicion that the s. 286.1(2) offence was occurring in a space defined with sufficient precision?

  (ii) If so, were the police entitled to offer the opportunity to commit child luring offences under ss. 172.1 and 172.2 of the Criminal Code?

[ 27 ] My reasons proceed as follows. First, I review the law of entrapment and consider how the bona fide inquiry prong accounts for the unique features of online spaces. Those features — in particular, that the Internet's boundaries are informational rather than geographical; that people behave differently online than they do in person; and that Internet use raises distinct rights concerns — call for greater attention to the functions and interactivity of online spaces when assessing whether that space was sufficiently precise to ground reasonable suspicion.

[ 28 ] Second, I apply that framework to this case and conclude, like the Court of Appeal, that the application judge erred in failing to consider circumstances beyond the number of people affected by the investigation. On a proper analysis, the police had reasonable suspicion that the criminal activity in question was occurring in a sufficiently precise space and the child luring offences were rationally connected and proportionate to the s. 286.1(2) offence. Project Raphael was thus a bona fide inquiry and Mr. Ramelson was not entrapped.

V. Legal Framework

A. The Entrapment Doctrine

[ 29 ] Whatever their utility in fighting crime, some police techniques are "unacceptable in a free society with strong notions of fairness, decency, and privacy" (Ahmad, at para. 16). Entrapment is one of them. It is not a traditional defence, but a form of abuse of process whose only remedy is a stay of proceedings. It may occur in two ways:

(a) the authorities provide a person with an opportunity to commit an offence without acting on a reasonable suspicion that this person is already engaged in criminal activity or pursuant to a bona fide inquiry;

(b) although having such a reasonable suspicion or acting in the course of a bona fide inquiry, they go beyond providing an opportunity and induce the commission of an offence.

(Mack, at pp. 964-65)

[ 30 ] As a form of abuse of process, the entrapment doctrine flows from courts' inherent jurisdiction to protect the justice system's integrity, a power necessary to preserve "the respect and support of the community", upon which the rule of law depends (R. v. Conway, [1989] 1 S.C.R. 1659, at p. 1667; see Mack, at p. 938). Like abuse of process, the entrapment doctrine censures state conduct that "violates our notions of 'fair play' and 'decency' and which shows blatant disregard for the qualities of humanness which all of us share" (Mack, at p. 940).

[ 31 ] Entrapment recognizes that "police involvement in the commission of a crime can bring the administration of justice into disrepute" (Ahmad, at para. 16). When the police offer opportunities to commit crimes without reasonable suspicion, or induce their commission, they may transgress several expectations: that the police will not intrude on privacy; that they will not randomly test the public's propensity to commit crimes, and still less manufacture them; that they, of all actors, will not act unlawfully for the purpose of entrapping others; and that they will not squander public resources on any of the above (Mack, at p. 958). Violating those expectations reflects poorly on law enforcement, but it may also diminish confidence in the justice system more generally.

[ 32 ] The remedy for entrapment is a stay of proceedings — the "most drastic remedy a criminal court can order" (R. v. Babos, 2014 SCC 16, [2014] 1 S.C.R. 309, at para. 30) — not because the accused is entitled to an acquittal, but because the Crown "is disentitled to a conviction" (R. v. Jewitt, [1985] 2 S.C.R. 128, at p. 148; Mack, at p. 944). A stay of proceedings ends prosecutions that infringe basic norms, marking courts' refusal to "condone or be seen to lend a stamp of approval to behaviour which transcends what our society perceives to be acceptable on the part of the state" (Mack, at p. 942).

[ 33 ] Yet law enforcement also serves an important public interest. The police must innovate if they are to match offenders' ingenuity. Some offences, too, are hard to investigate: whether because they are "consensual"; because they "victimize those who are reluctant or unable to report them"; or because they may "lead to such great harm that they must be actively prevented" (Ahmad, at para. 18). And so drastic a remedy as a stay of proceedings calls for some restraint. These realities entitle the police to "considerable latitude" in their investigations (Mack, at p. 917), such that a finding of entrapment should issue only in the "clearest of cases" (p. 976).

[ 34 ] The doctrine thus strives to balance competing imperatives: "The rule of law, and the need to protect privacy interests and personal freedom from state overreach . . ." on the one hand, and "the state's legitimate interest in investigating and prosecuting crime" on the other (Ahmad, at para. 22; see Mack, at pp. 941-42). Below, I consider what that balance requires in the context of police investigations into online spaces.

B. Bona Fide Inquiries and the Internet

(1) Overview

[ 35 ] The central issue on appeal is whether Project Raphael was a bona fide inquiry. This has two criteria: the police must have had (1) reasonable suspicion over a sufficiently precise space; and (2) a genuine purpose of investigating and repressing crime (Ahmad, at para. 20). Satisfying those criteria entitles the police to present "any person associated with the area with the opportunity to commit the particular offence" — even without individualized suspicion in the person investigated (R. v. Barnes, [1991] 1 S.C.R. 449, at p. 463 (emphasis in original)).

[ 36 ] Bona fide inquiries serve a useful purpose. In some cases, without knowing who might offend, the police may reasonably suspect that certain criminal activity is occurring within a given space. And, depending on the crime investigated, proactive methods may be necessary. The bona fide inquiry prong recognizes the legitimacy of such investigations, even as it seeks to confine them within careful limits.

[ 37 ] The Court has twice applied this prong of the entrapment doctrine. In Barnes, the police suspected that considerable drug trafficking was occurring in a six-block area of Vancouver's Granville Mall. The Court held that since the police's suspicions were reasonable, they were entitled, acting in good faith, to approach anybody associated with the space with an offer to purchase drugs. In Ahmad, the Court considered two separate appeals in which police suspected that two phone numbers were being used for dial-a-dope operations. But since their suspicion was based on anonymous, unverified and uncorroborated tips, the Court held that the police did not have reasonable suspicion of illegal activity for the phone number itself, and so were not acting under a bona fide inquiry.

[ 38 ] Barnes loomed large in the courts below. For the Court of Appeal, it was the "archetypical example" of a bona fide inquiry (para. 23), showing that such investigations can potentially target thousands of people (para. 79). And for the application judge in Jaffer, a companion case, "an analogy c[ould] readily be drawn", in the Internet context, "to the investigation of a geographical area in which specific crimes are known to be occurring to police" (A.R., Jaffer, vol. I, at p. 26).

[ 39 ] Yet the analogy calls for scrutiny. Physical spaces and actions are inherently limited in ways that virtual spaces and actions are not. The police in Barnes could have interacted with only so many people, approaching them one-by-one. And a physical space confines; it encloses people and things, and exhibits cycles of bustle and calm. Virtual spaces may escape those limits, inviting multitudes at all hours and distances, thus defying the boundaries we take for granted in the physical world. This explains why the Internet "provides fertile ground for sowing the seeds of unlawful conduct on a borderless scale" (R. v. Hamilton, 2005 SCC 47, [2005] 2 S.C.R. 432, at para. 30). And it explains why online police investigations may bear far-reaching implications for state surveillance, civil liberties and the rule of law.

[ 40 ] The space of the police investigation in Ahmad — a phone number — was inherently limited: functionally, it only allowed the police to contact the person who could answer the phone. Yet the Court foresaw that privacy would be a major concern in virtual investigations more generally. The "breadth of some virtual places . . ., the ease of remote access to a potentially large number of targets that technology provides law enforcement, and the increasing prominence of technology as a means by which individuals conduct their personal lives", it wrote, made state surveillance over virtual spaces qualitatively different than surveillance in public spaces (paras. 36-37). The nature of those differences falls to be further considered here, in the context of the Internet.

[ 41 ] This appeal, then, requires the Court to apply Ahmad, to further address how virtual and physical spaces differ, and to consider what those differences imply for the law. I look first at online spaces before considering the bona fide inquiry prong in more detail.

(2) Online Spaces

[ 42 ] Like any evolving technology, even an expansive definition of the Internet risks a quick obsolescence. And over time, the Internet has proven to be many things: social and anti-social, informative and mis-informative, and marked by clusters of hyperactivity and landscapes of inactivity. Still, some general features can be identified.

[ 43 ] While sustained by a vast physical infrastructure, the Internet, at least as most users experience it, is first and foremost a network of information and a means of connecting with others. That information is stored on servers but is accessible from an increasing array of devices in many physical locations. To say that individuals gather in online "spaces" means only that people have accessed shared information, wherever they happen to be geographically — which, today, may be from nearly anywhere.

[ 44 ] Freed of geographical constraints, online spaces permit unique experiences. They are permeable, allowing users to seamlessly traverse from one space to another. They are often interactive, facilitating that movement and encouraging users to express themselves and engage with content. They can also be coded to enable varying levels of supervision, regulation or control. The Internet can be manipulated in ways that physical spaces cannot.

[ 45 ] And its information comes in all types. Functionally, the Internet encompasses the most public and the most private human behaviour. It is the largest megaphone or billboard ever conceived, allowing publishers to connect with audiences far vaster than could ever physically congregate. Yet many millions also conduct private activities online, confident that their information — whether touching their work, social or personal lives — will remain as secure from general circulation as if they had transacted in person.

[ 46 ] With that range of behaviour comes a range of candour. Some online locations, like search engines, allow people to explore notions that they would be loath to air in public; others, like some forms of social media, allow users to dissimulate behind veneers of their choosing. Still others, like those dedicated to sexual activity, may encompass both poles. Online behaviour, in other words, may be radically transparent, radically disingenuous, or both. People do not always act online as they do in person.

[ 47 ] This, combined with its ubiquity, helps explain why the Internet raises "a host of new and challenging questions about privacy" (R. v. Spencer, 2014 SCC 43, [2014] 2 S.C.R. 212, at para. 1). Virtual police investigations may produce vast amounts of highly personal information, in contexts where people may be unusually uninhibited, engaged in forms of self-discovery, or seeking anonymity. The mere threat of state intrusion into those spaces may promote self-censorship, or the abstention from those spaces altogether, with costs to free expression and the exchange of ideas so essential in a vibrant democracy.

[ 48 ] Nor do privacy concerns end with online interactions. The Internet collects traces — information about a user's physical location, online activity, and more — in ways that in-person interactions typically do not (Spencer, at para. 46). This data collection often occurs without a user's awareness or consent, and those traces may persist indefinitely. They can spread with prodigious speed and reach, making it still more likely those traces will persist. And they can be compiled, dissected and analyzed to lend new insights into who we are as individuals or populations. As the rights and autonomy implications of those insights have become clearer, the divisions over how data is collected, protected and mobilized have in turn sharpened.

[ 49 ] Online spaces, in short, differ from physical spaces in at least three ways: the Internet, being informational rather than geographical, sheds many of the physical world's limitations in terms of scale and functions; people behave differently when online; and virtual spaces raise unique rights concerns. As Ahmad stated, online spaces are qualitatively different.

[ 50 ] In my view, the greatest consequence of these differences for bona fide inquiries is that the boundaries of an online "space" only tell part of the story. While intuitive, geographical analogies are imperfect. There is no simple way to compare a six-block area of a city to its online equivalent, except perhaps via those spaces' functions. And even then, similar functions may conceal deeper differences in experience. In an era when a single Tweet may attract more traffic than an entire mall, the parameters of a virtual space may be a poor proxy for the scope of a police investigation.

[ 51 ] To respect the entrapment doctrine's underlying balance of first principles, then, courts assessing whether an online police investigation was bona fide must pay close attention to the space's functions and interactivity — that is, to the permeability, interconnectedness, dynamism and other features that make the Internet a distinctive milieu for law enforcement. Even tailored online investigations may represent a broad and profound invasion into peoples' lives. Given the potential of online investigations to impact many more individuals than an equivalent investigation in a physical space, the nature of those impacts deserve scrutiny. How the police act on the Internet may matter as much or more as where they act.

(3) Bona Fide Inquiries and Reasonable Suspicion

[ 52 ] The question becomes how the test for bona fide inquiries applies in the context of virtual spaces; a question this Court first addressed in Ahmad. As noted, bona fide inquiries must satisfy two criteria before the police may offer an opportunity to commit an offence: the police must have (1) a reasonable suspicion over a sufficiently precise space; and (2) a genuine purpose of investigating and repressing crime (Ahmad, at para. 20). The second criterion was not raised as an issue in this case. The issue, instead, is whether the police had reasonable suspicion over a sufficiently precise space.

[ 53 ] Reasonable suspicion is itself a familiar standard, applicable in other contexts (Ahmad, at para. 30). It is not onerous; it requires only the reasonable possibility, not probability, that crime is occurring (R. v. Chehil, 2013 SCC 49, [2013] 3 S.C.R. 220, at para. 27). Yet it still subjects police actions to "exacting curial scrutiny", to ensure they were founded on objective evidence rather than on profiling, stereotyping or other improper grounds (Ahmad, at paras. 24-25). As an objective standard, it "protects everyone from random testing", whether they are tempted to commit crimes in the space or not (para. 27 (emphasis in original)).

[ 54 ] In its application, reasonable suspicion requires "a constellation of objectively discernible facts assessed against the totality of the circumstances" (R. v. Stairs, 2022 SCC 11, [2022] 1 S.C.R. 169, at para. 68). It is "fact-based, flexible, and grounded in common sense and practical, everyday experience" (Chehil, at para. 29). It eschews generalities; the suspicion must be "sufficiently particularized" (para. 30). And while it may be informed by an officer's training and experience, it must be concretely grounded; it cannot rest on hunches, intuition, or mere educated guesses (para. 47).

[ 55 ] The phrase "random virtue testing" has denoted police investigations that overstep the entrapment doctrine's limits (see Mack, at p. 956; Barnes, at p. 463). Some degree of "randomness" is inherent in bona fide inquiries: in approaching individuals without suspecting their personal involvement in crime, it is "inevitable" that such investigations "might unfortunately result in attracting into committing a crime someone who would not otherwise have had any involvement in criminal conduct" (Mack, at p. 956). In general, this risk only increases as a space is defined more broadly. And a broadly defined space could also cause an investigation to unduly burden the public's ability to "go about their daily lives without courting the risk that they will be subjected to the clandestine investigatory techniques of agents of the state" (Ahmad, at para. 39, citing Barnes, at p. 480, per McLachlin J., dissenting).

[ 56 ] The space where police reasonably suspect crime is occurring must therefore be "carefully delineate[d]" and "defined with sufficient precision", to "ensure the police have narrowed their scope so that the purview of their inquiry is no broader than the evidence allows" (Ahmad, at paras. 39 and 41). This is all the more critical in virtual spaces, which, lacking the constraints inherent in physical spaces, can be extremely vast in reach. While true of all spaces police may investigate, it is particularly important that online spaces are "defined narrowly and with precision" (para. 43).

[ 57 ] In Ahmad, the Court listed six factors that may illuminate the assessment of whether the police investigation was properly tailored: (1) the seriousness of the crime in question; (2) the time of day and the number of activities and persons who might be affected; (3) whether racial profiling, stereotyping or reliance on vulnerabilities played a part in the selection of the location; (4) the level of privacy expected in the area or space; (5) the importance of the virtual space to freedom of expression; and (6) the availability of other, less intrusive investigative techniques (para. 41).

[ 58 ] The Ahmad factors are contextual guides; they are neither exhaustive nor mandatory. And while they help in maintaining the entrapment doctrine's underlying balance between "the state's interest in investigating crime and the law's constraint against unwarranted intrusion into individuals' personal lives" (para. 63), the factors do not themselves require balancing. All six factors will not always be relevant or helpful — an otherwise overbroad police inquiry, for instance, is not saved by the fact that it did not involve profiling or stereotyping. But nor should one factor be allowed to overwhelm the inquiry. Even those convicted of the most serious offences may have been entrapped; conversely, those acting in the most private spaces will not necessarily be entrapped. The analysis is always contextual.

[ 59 ] The Court in Ahmad also made clear that the reasonable suspicion analysis turns not only on a space's physical characteristics, but on whether the police "have narrowed their scope so that the purview of their inquiry is no broader than the evidence allows" (para. 41). This necessarily engages with the nature of the police's suspicions, including the type and extent of the criminal activity suspected, and the nature of their investigation. Indeed, the factors expressly address the "seriousness of the crime in question" and an investigation's impacts on people and activities in the time and place in which it occurs (para. 41). This reflects the fact that reasonable suspicion of particular criminal activity, in bona fide inquiries, is not freestanding; it exists in relation to a particular space and justifies the state in offering opportunities to commit certain crimes to "any person associated with the area" (Barnes, at p. 463 (emphasis in original)). The space, the crimes and the nature of the investigation all influence the acceptable purview of the police's inquiry. The entire context, in short, determines whether the space of an investigation was sufficiently precise.

[ 60 ] Still, to date, the need for sufficient precision has played relatively little role in this Court's cases. The investigation in Barnes grew out of suspicion of drug trafficking in the six-block area of a city, and took place in that same broad space, a narrower investigation being "unrealistic" in the circumstances (p. 461). The dial-a-dope investigation in Ahmad, too, both originated in and occurred in the virtual space of a phone number, a space that, functionally, could hardly have been more tailored. And in inherently limited spaces like those, the scope of police actions can typically only extend so far, attenuating the risks that police actions will exceed what the evidence permits.

[ 61 ] But such scrutiny will often be crucial when the police investigate broader virtual spaces like a website, for at least two reasons.

[ 62 ] First, in permeable and interactive spaces like websites, which can be subdivided in various ways, the "space" of an inquiry will not necessarily be intuitive. Project Raphael, for instance, grew out of the police's suspicion that crime was occurring through user-created ads for the youngest sex workers posted in the escort subdirectory of Backpage; yet it was designed to only offer opportunities to those users who entered and responded to police-created ads, which were modeled after those user-created ads. The possibility of creating subspaces like this — such as postings, messages or hyperlinks within a broader website — suggests that the descriptions of online spaces may be more dynamic than those of physical spaces. And it suggests that the ways subspaces are embedded in, or relate to other online spaces may be critical for understanding how the space of an inquiry was "premised upon and tethered to reasonable suspicion" (Ahmad, at para. 20). The space — which both grounds reasonable suspicion and defines the inquiry's purview — should thus be "carefully delineate[d]" (para. 39). The area of an online police investigation calls for close attention.

[ 63 ] Second, whether an online space was sufficiently precise may turn as much on the space's functions and interactivity as it does on its parameters. As noted, the Internet may allow the police to reach many more people, in more targeted or invasive ways, than equivalent investigations into physical spaces. For instance, the investigation in Barnes (where the police acted as in-person buyers) was unlikely to generate the flood of responses that overwhelmed the YRP in Project Raphael's first "phase" (where the police acted as online sellers). Nor, without great efforts, could it have produced the same record of users' personal information, including phone numbers, which was produced automatically through texting. And a website may allow the police to "fabricate identities and 'pose' as others to a degree that would not be possible in a public space like the Granville Mall" (Ahmad, at para. 37) — for example by veiling their identities through images or other digital representations.

[ 64 ] Such functions may require tailoring the location of an online inquiry to a degree that was unnecessary in either Barnes or Ahmad, where the space's scope and functions were already limited. They may, for example, require the police to focus on more carefully delineated spaces and target their opportunities to particular subspaces or to particular ways in which users engage with the space. This is especially true in places frequented by vulnerable groups, such as racial, religious or sexual minorities, or in spaces whose use carries important rights implications, where the need for precision is particularly critical (see Ahmad, at paras. 169-70, per Moldaver J., dissenting; D. M. Tanovich, "Rethinking the Bona Fides of Entrapment" (2011), 43 U.B.C. L. Rev. 417, at p. 432; K. Roach, "Entrapment and Equality in Terrorism Prosecutions: A Comparative Examination of North American and European Approaches" (2011), 80 Miss. L.J. 1455, at pp. 1474-75; R. v. Stack, 2022 ONCA 413 (raising but not deciding whether the accused was entrapped on Grindr, a "dating and hookup" app for men seeking men)).

[ 65 ] A space's functions may also, in some cases, call for attention to how the space facilitates or inhibits data collection. The Internet's ability to collect traces may cut in different directions. Because the accused bears the burden of proof in entrapment applications, a lack of data about the scope of an investigation — such as how many people it contacted, and in what ways — could raise fairness concerns. Yet indiscriminate data collection, without appropriate safeguards, would raise clear privacy concerns. Generally speaking, and insofar as an online space permits, police should consider whether to collect and retain data relevant to understanding an investigation's scope, while taking care to do so in a way that minimizes the costs to privacy.

[ 66 ] The Ahmad factors, again, may assist. An online space's functions and interactivity can greatly influence factors like the "number of activities and persons who might be affected", the "level of privacy expected", the "importance of the virtual space to freedom of expression" or whether the police's methods were sufficiently narrow as compared to "other, less intrusive investigative techniques" (para. 41). These, alongside other concerns, may be critical in assessing whether the space was "defined narrowly and with precision" (para. 43).

[ 67 ] None of this is to say that broad online spaces cannot be sufficiently precise. In Ahmad, the Court explained that "entire websites or social media platforms will rarely, if ever, be sufficiently particularized to support reasonable suspicion" (para. 43). Indeed, multi-functional virtual spaces — perhaps typical of much Internet use — will usually be too broad to support reasonable suspicion. But in other spaces, the criminality may be so pervasive that it supports a reasonable suspicion over the entire area (see R. v. Nelson, 2021 BCCA 192, at paras. 13-19 (trafficking fentanyl and carfentanil on the dark web)). The circumference of a sufficiently precise online location may vary.

[ 68 ] In sum, the Internet's unique features are inescapable in assessing whether the location is sufficiently precise to ground reasonable suspicion. Online spaces are qualitatively different (Ahmad, at para. 37) — and those differences must be considered.

C. Rationally Connected and Proportionate Opportunities

[ 69 ] A second doctrinal issue arises from the facts of this case. The police suspected that the offence of communicating to obtain sexual services from a minor (s. 286.1(2)) was occurring on Backpage. But the Backpage website prevented the police from directly advertising sexual services by someone under the age of 18; instead, they had to reveal a younger age in conversation. Yet the age they revealed — 14 years — exposed Mr. Ramelson to more serious offences of child luring a person under 16. This raises the further question: what type of link must exist between the offence suspected and the offence offered?

[ 70 ] Citing Mack, the Court of Appeal concluded that the child luring under 16 offences were not "totally unrelated" or "much more serious" offences than the s. 286.1(2) offence, which concerned a person under 18 years old (para. 89). Yet the phrases "totally unrelated" and "much more serious" only served in Mack to identify cases that clearly met the standard; they were not the standard itself. That standard remains the one set in Mack: that the crime that police offer must be rationally connected and proportionate to the offence they suspect is occurring.

[ 71 ] In the companion appeal Haniffa, Mr. Haniffa argues that bona fide inquiries, being inherently broader than investigations into individuals, ought to be restricted to offering opportunities to commit the same offences that the police suspect are occurring. But given the restrictions on bona fide inquiries that I have discussed above, such a distinction is, in my view, unnecessary. And it could unduly hinder the police's ability to investigate certain offences. I would therefore reject it.

VI. Application

[ 72 ] Mr. Ramelson's appeal raises two issues of application: (1) whether the police had reasonable suspicion over a sufficiently precise space; and (2) if so, whether the child luring offences he was charged with were rationally connected and proportionate to the s. 286.1(2) offence.

[ 73 ] Like the Court of Appeal (at paras. 96-97), I conclude that the application judge erred by failing to properly consider the broader context beyond the number of people affected by the investigation. On the proper analysis, I conclude that Project Raphael was bona fide and Mr. Ramelson was not entrapped. I consider each issue in turn.

A. Did the Police Have Reasonable Suspicion in a Sufficiently Precise Space?

[ 74 ] The police suspected that the s. 286.1(2) offence (communicating to obtain sexual services from a minor) was occurring within ads posted in the escort subdirectory of the York Region Backpage whose content suggested extreme youth. I agree with the Court of Appeal that the application judge erred in finding that space was insufficiently precise to ground reasonable suspicion.

[ 75 ] Whether the police had reasonable suspicion turns on Insp. Truong's testimony. In some cases, a single officer's testimony may fail to support reasonable suspicion, particularly if too closely tied to their own "personal conclusions" rather than objective fact (Ahmad, at para. 24 (emphasis deleted)). But here, it was sufficient. Insp. Truong's testimony was grounded in his direct and indirect experiences in law enforcement: the YRP had identified 85 juvenile sex workers between 2011 and 2016; the 2013 seller-side initiative, which located 9 juveniles in 2 weeks and found that those investigated entered the industry at 14.8 years old on average; he had participated in many conferences and conversations in his professional circles; and he had interviewed hundreds of sex workers in his several years specializing in sex work-related crimes — all of which pointed to a pervasive problem stemming from a particular type of online ad on Backpage.

[ 76 ] This, in my view, amply showed the reasonable possibility that the s. 286.1(2) offence was occurring in the space. Indeed, it suggested that the offence was occurring regularly. If the YRP were to address offences related to juvenile sex work, ads in the York Region escort subdirectory of Backpage for the youngest sex workers were places to do so.

[ 77 ] The question becomes whether the officers' reasonable suspicion related to a space that was sufficiently precise. Mr. Ramelson submits that the application judge was correct to find that Project Raphael's scope was overbroad, since the investigation extended to an "entire website". In stressing the need to click on and respond to the ads, and downplaying the need for closer targeting to those seeking juveniles, the Court of Appeal's approach will "result in almost any website or social media platform being defined as narrow, even if in practice it is extremely wide" (A.F., at para. 47). And the court further discounted the number of people affected by the investigation, which was significant. The Crown responds that the space was sufficiently precise: the escort section was already limited by nature, and the information in the ads further filtered out potential respondents, to the extent that Backpage's parameters allowed. The Court of Appeal's review of the Ahmad factors — which emphasized the seriousness of the crime and the difficulty investigating it — reflected the correct approach.

[ 78 ] I agree that the application judge erred. While correctly stressing the number of people affected by the investigation, he failed to properly consider the entire context — in particular, the seriousness of the crimes and the difficulty investigating them via alternative techniques. Like the Court of Appeal, a review of the full context leads me to conclude that the online space in which Project Raphael offered opportunities was defined with sufficient precision to ground the police's reasonable suspicion. I begin with the virtual space's definition, which must be carefully delineated, including, as I have explained, with a view to the space's functions and interactivity.

[ 79 ] At its most general, the YRP's suspicions arose within the York Region escort subdirectory of the Backpage website. There was, to all appearances, an active hub of crime; specific posts advertising the youngest sex workers appeared daily, and unremittingly, within that broader space. And in designing Project Raphael, the YRP sought to replicate those posts, aiming to attract those who were either actively seeking a juvenile or were at least indifferent to encountering one. The space — grounding reasonable suspicion and defining the investigation's purview — was the particular type of ads within the York Region escort subdirectory of Backpage that emphasized the sex worker's extreme youth.

[ 80 ] To be sure, that space differed markedly from the physical spaces of traditional investigations. First, at a granular level, the location where the police's suspicions first arose (the user-created ads in that subdirectory) was not identical to the location where the police later offered their opportunities (the police-created ads in the escort subdirectory); indeed, the police-posted ads did not exist until the police created them. Those specific ads, therefore, could not generate the reasonable suspicion upon which Project Raphael was grounded: the reasonable suspicion standard "insists on an objective assessment of the information the police actually had" (Ahmad, at para. 29). That assessment turns instead on the police's suspicions over a certain type of ad for the youngest sex workers that was arising continually, and without the police's involvement, within the broader York Region escort subdirectory of Backpage. The relationship between those user-created ads and the police-created ads — both being the same type of ads within the same subdirectory — is thus integral to the space's definition: it explains how the police-created ads could be "premised upon and tethered to reasonable suspicion" (para. 20).

[ 81 ] Second, Project Raphael illustrates how an investigation's design may shape — and tailor — the nature of an online space according to its functions and interactivity. In modeling Project Raphael on common, independently posted ads for the youngest sex workers, the police consciously designed their ads to be functionally analogous to pre-existing ads, and to draw in a similar audience. And in listing a phone number for users to text message — a step made necessary by their functional inability to communicate through the website with those who merely clicked on the ad — the police shaped how they would interact with users and, ultimately, how they would offer opportunities to commit the offences. The space's functions and interactivity, in other words, permitted the police to design Project Raphael in a way that narrowed the investigation's scope.

[ 82 ] I therefore disagree with Mr. Ramelson that the police's investigation extended to an "entire website". Even the escort subdirectory as a whole was not an entire website. And users within that broader space had to take interactive steps to locate the youngest sex workers. Unlike in Barnes, where anybody within the six-block area could have been approached by the police, Backpage users would only encounter the police if they clicked on particular ads, entering police-designed subspaces, and then messaged the listed phone number. While anyone within the escort subdirectory could view the links to those postings, Project Raphael was not designed to interact with everyone within that broader space, let alone within the Backpage website as a whole.

[ 83 ] As noted, the ads themselves were designed, in word and image, to signal youth. In Mr. Ramelson's case, the ad called the sex worker a "Tight Brand NEW girl . . . who is sexy and YOUNG with a tight body", with a "YOUNG FRIEND if your interested too". The photo featured a faceless woman wearing a t-shirt that bore the name of a local high school. And her stated age was the minimum compatible with the space's parameters. While online information cannot always be taken at face value, and while other steps could have been taken — in one of the related appeals, Jaffer, the police included the number "16" in an email address, for instance — these were clear hints that the sex worker might have been underage. This further narrowed the space within which the police offered their opportunities.

[ 84 ] Mr. Ramelson was among those who took things still further: by not only browsing the escort subdirectory, and clicking on the hint-adorned ad, but by messaging the affiliated phone number and arranging an encounter — all before the police mentioned the sex worker's age. It was at this point — when the police mentioned the sex worker's age — that they provided him with the opportunity to commit the offences under ss. 286.1(2), 172.1 and 172.2 (see Ahmad, at paras. 63-64; R. v. Ghotra, 2020 ONCA 373, 455 D.L.R. (4th) 586, at paras. 30-31, aff'd 2021 SCC 12, [2021] 1 S.C.R. 521). By agreeing to proceed with the transaction, all the elements of the offences were satisfied (R. v. Levigne, 2010 SCC 25, [2010] 2 S.C.R. 3, at para. 23).

[ 85 ] These features limited Project Raphael's purview. Still, being an online investigation, the number of people it affected was potentially far greater than an analogous investigation in a physical space, even with users having to take steps to encounter the police. Because the police never kept statistics, there are no precise numbers. But most who responded to the police's ads, it appears, immediately disengaged once they were told the sex worker's age (C.A. reasons, at paras. 141-42). This supports an inference that Project Raphael affected many who would not have offended independently. For the application judge, this was a crucial concern. And I agree that that this raises legitimate concerns.

[ 86 ] Much of Project Raphael's breadth lay in the nature of the police's offer. The point is again made clear through a comparison with Barnes. The police there approached people individually, seeking those with easy access to drugs who were willing sellers. This necessarily limited the number of people who could actually take advantage of the opportunity. Here, by contrast, the investigation occurred online, through ads potentially available to everyone in the escort subdirectory, and sought those who, already intent on a sexual transaction, were willing buyers. This made it more likely that users would avail themselves of the police's opportunity, and more likely that the investigation would ensnare those who would not have committed the offences on their own.

[ 87 ] A related issue was Project Raphael's conscious targeting of those who were only indifferent about the sex worker's age. This troubled the application judge: he noted that it captured individuals who were not seeking to commit the suspected offence. By definition, such people may not have offended absent the police's offer, suggesting that their inclusion in Project Raphael may have manufactured crimes that would not have occurred on their own — one of the key risks targeted by the entrapment doctrine (Barnes, at p. 459; Ahmad, at para. 28).

[ 88 ] In this specific context, however, I agree with the Court of Appeal that the "merely indifferent" were legitimate targets of the investigation (para. 124). As inchoate offences, the s. 286.1(2) and child luring offences target those who, knowing the person's age, or being willfully blind to it, decide to pursue sexual activity anyway — regardless of whether they initially sought a juvenile or whether they actually carry out the agreement. Aiming to "close the cyberspace door" before things proceed any further (R. v. Legare, 2009 SCC 56, [2009] 3 S.C.R. 551, at para. 25), the offences embody a particularly proactive approach to crime, criminalizing mere agreements to proceed in the face of certain facts. Capturing the "merely indifferent" only reflects the nature of the offence. It reflects Parliament's own judgment about when such conversations cross a line, and suggests there is a legitimate law enforcement interest in the police intervening at a relatively early stage.

[ 89 ] Another key concern was privacy. I agree with the Court of Appeal that Project Raphael "intruded upon an intensely personal privacy interest" (para. 135). The text message communications that followed for users who clicked on the ads and contacted the phone number disclosed intimate details about those users' sexual preferences, details they would naturally expect would remain confidential from the state.

[ 90 ] Privacy and free expression interests are often interrelated, since a loss of privacy often precedes self-censorship. The Court of Appeal also concluded that, being illegal and falling outside traditional categories of expression valued in a democratic society, buyers' expression in this case carried little weight (para. 136). For similar reasons, it concluded that Project Raphael did not intrude on other public activities of value (paras. 127-29). But although crime was apparently pervasive within the types of ads targeted by the investigation, communications in those spaces were not necessarily illegal — a person may communicate with a sex worker without intending to hire them for sex. And the police's presence in an online space may chill any of the expression in that space, legal or illegal. Still, I agree that Project Raphael did not intrude on public activities of great value and the privacy interests in this case were greatly attenuated.

[ 91 ] Most importantly, these factors were only part of the context in which police offered the opportunities in Project Raphael. The nature of the offences and the lack of alternative investigative techniques were key factors in this case; yet ones the application judge mentioned only briefly (second ruling on entrapment, at para. 22).

[ 92 ] The offences were not only apparently prevalent; they were serious and difficult to investigate except through proactive methods. All involved communications to arrange a sexual encounter with a minor, a "very real threat", for which Parliament, recognizing their gravity, has criminalized preparatory actions to the intended conduct itself (R. v. Morrison, 2019 SCC 15, [2019] 2 S.C.R. 3, at paras. 39-40). The difficulty investigating "consensual" crimes of this type has long been recognized, even before those crimes moved online (Kirzner v. The Queen, [1978] 2 S.C.R. 487, at pp. 492-93; Amato v. The Queen, [1982] 2 S.C.R. 418, at p. 457, per Estey J., dissenting, but not on this point; Mack, at pp. 916-17; Ahmad, at para. 18). And sting operations have become "an important tool — if not the most important tool — available to the police in detecting offenders who target children and preventing them from doing actual harm to children" (R. v. Friesen, 2020 SCC 9, at para. 94). Given the "considerable latitude" police are owed in their investigations (Mack, at p. 917), sting operations like Project Raphael should not be foreclosed lightly.

[ 93 ] In all the circumstances, I would conclude that ads posted in Backpage's escort subdirectory emphasizing the sex worker's extreme youth were a sufficiently precise space to ground reasonable suspicion. Project Raphael carefully tailored the ads in which police provided the opportunity to commit the offences. The broader escort subdirectory was designed to facilitate sex-work related crimes, limiting the targeted audience. Like the user-created ads, the police-created ads indicated extreme youth, including showing the person wearing a t-shirt with the logo of a local high school. Users had to interact with those ads by text message to encounter the police. This was a serious inchoate offence involving juveniles. And it is unclear what alternative methods of investigation the police might have employed — the limitations of a seller-side investigation having already been demonstrated. Although the investigation impacted many individuals, in context, "the purview of the police inquiry [went] no broader than the evidence allow[ed]" (Ahmad, at para. 41).

B. Were the Offences Rationally Connected and Proportionate?

[ 94 ] Mr. Ramelson's final argument is that even if the police had reasonable suspicion that the s. 286.1(2) offence was being committed on the escort subdirectory of Backpage, it did not allow them to offer the opportunity to commit other, more serious offences under ss. 172.1(1)(b) and 172.2(1)(b). In other words, having reasonable suspicion that users were committing offences in respect of those under 18 did not allow them to offer opportunities in respect of those under 16. That "bait-and-switch" exposed Mr. Ramelson to a much longer sentence, for offences that were too remote from the police's suspicion.

[ 95 ] I disagree. While the police indeed lacked reasonable suspicion over the child luring offences, they were nonetheless rationally connected and proportionate to the s. 286.1(2) offence.

[ 96 ] I note first that the question only arises because of a lack of specificity in the evidence grounding the police's reasonable suspicion — while the police had reason to suspect that sex workers under 16 were active on Backpage, there was too little evidence that buyers were seeking them specifically. As the Court of Appeal correctly noted, the focus had to be on the buyers' intentions (para. 73).

[ 97 ] Some basic features of the offences suggest they are rationally connected. For one, they capture similar conduct, criminalizing telecommunications for the purpose of committing a sexual offence with a person under a specified age. The Court of Appeal noted their shared elements, explaining that the police would inevitably offer the opportunity to commit child luring in the course of offering the opportunity to commit the s. 286.1(2) offence (para. 86).

[ 98 ] The difference in age is concerning. Yet it rested on an evidentiary basis: those sex workers identified in the YRP's 2013 seller-side investigation, for instance, first started selling sex at age 14.8 on average. Even small differences in age can, of course, make a big difference in sexual offences. But in light of the facts, it is unsurprising, in my view, that their investigation into juvenile sex work would target buyers looking for those under 16.

[ 99 ] In my view, offering the opportunity to commit the child luring offences was not disproportionate. To use this Court's language in Mack, whether child luring is a "much more serious" offence than the s. 286.1(2) offence might be an indicator of proportionality. Granted, luring a person under the age of 16 is more serious and "subject to a longer period of incarceration" than committing the s. 286.1(2) offence in respect of a person under 18 (C.A. reasons, at para. 89). In some cases, differences in seriousness and punishment will go too far. Here, however, luring a person under 16 is not a disproportionately more serious crime than communicating to obtain sexual services from a minor. With a gap of six months separating the offences' mandatory minimum sentences, they remain comparable.

[ 100 ] I would therefore not accede to Mr. Ramelson's argument. The offences were rationally connected and proportionate, such that the police could offer opportunities to commit them given their reasonable suspicion in the s. 286.1(2) offence.

VII. Conclusion

[ 101 ] For these reasons, I conclude that the police had reasonable suspicion in a space defined with sufficient precision, and the offences offered were rationally connected and proportionate to each other. The police's genuine law enforcement purpose is not at issue. Project Raphael was thus a bona fide inquiry and Mr. Ramelson was not entrapped. I would dismiss the appeal.

Appeal dismissed.

Solicitors for the appellant: Hicks Adams, Toronto.

Solicitor for the respondent: Attorney General of Ontario, Crown Law Office — Criminal, Toronto.

Solicitor for the intervener the Director of Public Prosecutions: Public Prosecution Service of Canada, Toronto.

Solicitors for the intervener the Criminal Lawyers' Association of Ontario: Brauti Thorning, Toronto.

Solicitors for the intervener the British Columbia Civil Liberties Association: Stockwoods, Toronto.

Solicitors for the intervener the Canadian Civil Liberties Association: Paliare Roland Rosenberg Rothstein, Toronto.