WARNING

The President of the panel hearing this appeal directs that the following should be attached to the file:

An order restricting publication in this proceeding under ss. 486.4 or 486.6 of the Criminal Code shall continue. These sections of the Criminal Code provide:

486.4 (1) Subject to subsection (2), the presiding judge or justice may make an order directing that any information that could identify the victim or a witness shall not be published in any document or broadcast or transmitted in any way, in proceedings in respect of

(a) any of the following offences:

(i) an offence under section 151, 152, 153, 153.1, 155, 160, 162, 162.1, 163.1, 170, 171, 171.1, 172, 172.1, 172.2, 173, 213, 271, 272, 273, 279.01, 279.011, 279.02, 279.03, 280, 281, 286.1, 286.2, 286.3, 346 or 347, or

(ii) any offence under this Act, as it read from time to time before the day on which this subparagraph comes into force, if the conduct alleged would be an offence referred to in subparagraph (i) if it occurred on or after that day; or

(b) two or more offences being dealt with in the same proceeding, at least one of which is an offence referred to in paragraph (a).

(2) In proceedings in respect of the offences referred to in paragraph (1)(a) or (b), the presiding judge or justice shall

(a) as soon as feasible, inform any witness under the age of 18 years and the victim of the right to make an application for the order;

(b) on application made by the victim, the prosecutor or any such witness, make the order; and

(c) if an order is made, as soon as feasible, inform the witnesses and the victim who are the subject of that order of its existence and of their right to apply to revoke or vary it.

(2.1) Subject to subsection (2.2), in proceedings in respect of an offence other than an offence referred to in subsection (1), if the victim is under the age of 18 years, the presiding judge or justice may make an order directing that any information that could identify the victim shall not be published in any document or broadcast or transmitted in any way.

(2.2) In proceedings in respect of an offence other than an offence referred to in subsection (1), if the victim is under the age of 18 years, the presiding judge or justice shall

(a) as soon as feasible, inform the victim of their right to make an application for the order;

(b) on application of the victim or the prosecutor, make the order; and

(c) if an order is made, as soon as feasible, inform the victim of the existence of the order and of their right to apply to revoke or vary it.

(3) In proceedings in respect of an offence under section 163.1, a judge or justice shall make an order directing that any information that could identify a witness who is under the age of eighteen years, or any person who is the subject of a representation, written material or a recording that constitutes child pornography within the meaning of that section, shall not be published in any document or broadcast or transmitted in any way.

(3.1) If the prosecutor makes an application for an order under paragraph (2)(b) or (2.2)(b), the presiding judge or justice shall

(a) if the victim or witness is present, inquire of the victim or witness if they wish to be the subject of the order;

(b) if the victim or witness is not present, inquire of the prosecutor if, before the application was made, they determined if the victim or witness wishes to be the subject of the order; and

(c) in any event, advise the prosecutor of their duty under subsection (3.2).

(3.2) If the prosecutor makes the application, they shall, as soon as feasible after the presiding judge or justice makes the order, inform the judge or justice that they have

(a) informed the witnesses and the victim who are the subject of the order of its existence;

(b) determined whether they wish to be the subject of the order; and

(4) An order made under this section does not apply in either of the following circumstances:

(a) the disclosure of information is made in the course of the administration of justice when the purpose of the disclosure is not one of making the information known in the community; or

(b) the disclosure of information is made by a person who is the subject of the order and is about that person and their particulars, in any forum and for any purpose, and they did not intentionally or recklessly reveal the identity of or reveal particulars likely to identify any other person whose identity is protected by an order prohibiting the publication in any document or the broadcasting or transmission in any way of information that could identify that other person.

(5) An order made under this section does not apply in respect of the disclosure of information by the victim or witness when it is not the purpose of the disclosure to make the information known to the public, including when the disclosure is made to a legal professional, a health care professional or a person in a relationship of trust with the victim or witness.

486.6 (1) Every person who fails to comply with an order made under any of subsections 486.4(1) to (3) or subsection 486.5(1) or (2) is guilty of an offence punishable on summary conviction.

(1.1) A prosecutor shall not commence or continue a prosecution against a person who is the subject of the order unless, in the opinion of the prosecutor,

(a) the person knowingly failed to comply with the order;

(b) the privacy interests of another person who is the subject of any order prohibiting the publication in any document or the broadcasting or transmission in any way of information that could identify that person have been compromised; and

(c) a warning to the individual is not appropriate.

(2) For greater certainty, an order referred to in subsection (1) applies to prohibit, in relation to proceedings taken against any person who fails to comply with the order, the publication in any document or the broadcasting or transmission in any way of information that could identify a victim, witness or justice system participant whose identity is protected by the order.

COURT OF APPEAL FOR ONTARIO

CITATION: R. v. Pike, 2024 ONCA 608

DATE: 20240809

DOCKET: C70656, COA-23-CR-0023 & COA-23-CR-0638

Tulloch C.J.O., George and Monahan JJ.A.

DOCKET: C70656

BETWEEN

His Majesty the King

Appellant

and

Jeremy Pike

Respondent

DOCKET: COA-23-CR-0023 & COA-23-CR-0638

AND BETWEEN

His Majesty the King

Appellant (COA-23-CR-0023) /

Respondent (COA-23-CR-0638)

and

David Scott

Respondent (COA-23-CR-0023) /

Appellant (COA-23-CR-0638)

Michael Fawcett and Rebecca De Filippis, for the appellant (C70656 & COA-23-CR-0638) / respondent (COA-23-CR-0023), His Majesty the King

Daisy McCabe-Lokos, for the respondent, Jeremy Pike (C70656)

Naomi Lutes, for the appellant (COA-23-CR-0023) / respondent (COA-23-CR-0638), David Scott

Jennifer Conroy and Ian Bell, for the intervener/appellant Public Prosecution Service of Canada (C70656)

Samara Secter and Jocelyn Rempel, for the intervener Canadian Civil Liberties Association (C70656 & COA-23-CR-0023)

Spencer Bass, for the intervener Criminal Lawyers Association (Ontario) (C70656 & COA-23-CR-0023)

Heard: April 15-16, 2024

On appeal from the acquittals entered by Justice David E. Harris of the Superior Court of Justice on April 19, 2022, sitting without a jury (C70656).

On appeal from the convictions entered on October 13, 2022, with reasons reported at 2022 ONSC 5823, and from the sentence imposed on May 18, 2023, with reasons reported at 2023 ONSC 3023, by Justice David E. Harris of the Superior Court of Justice, sitting without a jury (COA-23-CR-0023 & COA-23-CR-0638).

Tulloch C.J.O.:

A. OVERVIEW

[1] Millions of Canadians travel internationally each year with their personal computers, tablets, and smartphones. The contents of these devices attract some of the strongest privacy interests known to law because they are a window into their users’ lifestyles, beliefs, interests, desires, relationships, finances, health, and much more. But one of Canada’s border laws, namely s. 99(1)(a) of the Customs Act, R.S.C. 1985, c. 1 (2nd Supp.), authorizes border officers to search those devices without any reasonable basis. Two travellers, Jeremy Pike and David Scott, whose devices were searched under this law, and who were consequentially charged with possessing and importing child pornography, challenged the constitutionality of the law. The principal issue in these appeals is whether s. 99(1)(a) of the Customs Act is constitutional.

[2] The answer is clear: It is not. Section 8 of our Charter of Rights and Freedoms guarantees everyone the right to be secure against unreasonable searches. I agree with the trial judge that s. 99(1)(a) of the Customs Act offends this basic guarantee because it authorizes border officers to search some of the most private information imaginable on the lowest possible standard to justify a search, namely that in the border officers’ own minds, they were sincerely trying to find evidence of border law violations. While sincerity is a good start, it is just not enough. Our Charter requires more because Canada’s border control interests temper but do not eliminate or gut its protections. A reasonable search in this context requires a reasonable suspicion, meaning objective facts that establish a reasonable possibility that officers will find evidence of border law violations on the device. The Crown in this matter has not justified the law’s failure to require those objective facts prior to conducting the searches, which the Customs Act requires for other highly intrusive searches and which the Canada Border Services Agency admits would not jeopardize its mandate.

[3] However, the unconstitutionality of this law does not entitle Mr. Pike and Mr. Scott to acquittals on the serious crimes against children with which they were charged.

[4] The trial judge correctly determined that a border officer breached Mr. Pike’s right to be secure from unreasonable searches and right to counsel when he detained him, asked him to provide the passwords to his digital devices and searched those devices without advising him of his right to counsel. However, the trial judge incorrectly excluded the evidence found on the devices because he failed to properly consider the border officer’s good faith reliance on the existing law when conducting the searches, Mr. Pike’s reduced expectation of privacy in his devices at the border, the moderate length of the delay in providing the right to counsel, and the interests of society in trying these serious charges on their merits. The trial judge’s decision to exclude the child pornography found on Mr. Pike’s devices, and acquit him, cannot stand. I would thus allow the Crown’s appeal against Mr. Pike’s acquittal, admit the evidence of child pornography found on his devices, and order a new trial on the child pornography charges.

[5] As for Mr. Scott, I agree with the trial judge that the files the border officers found on his devices are admissible because the border officers reasonably relied on the law’s validity. I would accordingly dismiss his appeal from conviction.

[6] The 23-month conditional sentence of house arrest that the trial judge imposed on Mr. Scott for importing the recordings of the sexual abuse and exploitation of children that he possessed, which the Crown seeks leave to appeal, also requires reconsideration due to the deleterious long-term impact of this type of offence on society at large, and on children in particular.

[7] These offences against children are abhorrent in nature and have lasting and widespread societal impact. By possessing recordings depicting the sexual abuse and exploitation of children over the course of three decades, Mr. Scott sexually abused children by invading their privacy, wounding their dignity, and causing them serious emotional harm. He shares culpability with those who produced the images depicting the victimization of children that he possessed on his digital devices. The offences for which he was charged normalize and increase the demand for further child abuse and exploitation. They increase the risk of sexual abuse of children and perpetuate demeaning messages that children are property or objects and not people. A child-centered approach to sentencing requires judges to consider child victims and the wrongs and harms that people who possess child pornography inflict on them, to reject myths that minimize the perpetrator’s responsibility and, finally, to apply a denunciatory sentencing range that reflects the abhorrent and harmful nature of these offences and their long-term negative impacts on children.

[8] Considering this guidance, the trial judge erred by imposing a conditional sentence of house arrest. Mr. Scott treated the children he sexually abused as property and objects to exploit and to add to his so-called “collection.” The dignity he denied to those children belies his attempt upon arrest to minimize his crime as a mere harmless fantasy. While the trial judge generally appreciated the seriousness of Mr. Scott’s crime, he overlooked the significant number of children Mr. Scott victimized and the three-decade duration of his offending. The appropriate sentence for Mr. Scott’s serious sexual offence against children should be three years imprisonment in the penitentiary, not 23 months’ house arrest.

[9] While the trial judge should have imposed a three-year sentence on Mr. Scott, I would decline to incarcerate him now because he has already served more than half of his 23-month house arrest sentence in the interim. He would likely have received parole by now if the trial judge had imposed a three-year sentence, and this decision’s condemnation of possessing child pornography and guidance for future cases is sufficient to denounce and deter that crime. Ordinarily, this court would allow the Crown’s appeal and impose the appropriate sentence to clarify what sentence the trial judge should have imposed, and then stay its execution. But doing so here would shorten Mr. Scott’s sentence by releasing him from the rest of his term of house arrest. I would accordingly grant the Crown leave to appeal the sentence but dismiss the sentence appeal to prevent that unfair result and ensure that Mr. Scott serves the rest of his house arrest sentence.

B. background

(1) Canada’s Border Law Framework

[10] In 1985, Parliament passed the Customs Act to regulate and control the movement of people and goods across its international border: Martineau v. M.N.R., 2004 SCC 81, [2004] 3 S.C.R. 737, at para. 25. The Act helps advance Canada’s national security, public health, crime control, immigration, tax collection, and other border policy objectives: R. v. Jones (2006), 2006 CanLII 28086 (ON CA), 81 O.R. (3d) 481 (C.A.), at para. 31 (“Jones (2006)”).

[11] Parliament tasked the Canada Border Services Agency (the “Agency”) via subsection 5(1)(a) of the Canada Border Services Agency Act, S.C. 2005, c. 38, with administering and enforcing the Customs Act and other border legislation. The Agency processes tens of millions of travellers, vehicles, and goods entering Canada each year. It is responsible for preventing the entry of unauthorized people into Canada and ensuring that travellers and importers pay taxes on lawful goods that they bring into Canada. Because not all goods brought into Canada are lawful, Parliament also charged the Agency with preventing the entry of prohibited goods: R. v. Simmons, 1988 CanLII 12 (SCC), [1988] 2 S.C.R. 495, at pp. 527-528. For example, subsections 159(1) and 160(1) of the Customs Act make it a penal offence punishable by up to 5 years in prison to import prohibited goods, including child pornography or illegal drugs. This conduct is also a crime under subsection 163.1(3) of the Criminal Code, R.S.C. 1985, c. C-46, and section 6 of the Controlled Drugs and Substances Act, S.C. 1996, c. 19. Both criminal statutes trigger even higher maximum sentences for importing, except for the least serious drugs.

[12] Parliament granted the Agency extensive powers to discharge its mandate and imposed corresponding obligations on travellers and importers to allow the Agency to conduct inspections to ensure compliance with the Customs Act. Agency officers have numerous powers to search people and goods, as outlined in sections 98-99.5 of the Act. People entering or importing goods into Canada, in turn, must present themselves to an Agency officer, report all goods they are importing into Canada, truthfully answer the officer’s questions, and make goods available for examination upon request, pursuant to sections 11-13 of the Act. All travellers complete primary processing and answer basic customs and immigration questions by either completing a physical declaration card and interacting with an Agency officer, or by using kiosk technology. Agency officers also have discretion to refer travellers to secondary examination, where officers can inspect goods and further question travellers to uncover non-compliance with the Customs Act.

[13] Section 2 of the Act defines “goods” to include “conveyances, animals and any document in any form.” The Agency has long taken the position that the electronic files which are contained within digital devices are documents and, thus, goods, and the Agency accordingly relies on the goods inspection powers the Act grants it to search the contents of digital devices. As summarized in R. v. Canfield, 2020 ABCA 383, 395 C.C.C. (3d) 483, at para. 69, leave to appeal refused, [2020] S.C.C.A. No. 367, most courts have agreed with this position.

[14] Many of the powers Parliament granted the Agency require its officers to reasonably suspect a border law violation before searching people and goods. For instance, subsections 98(1) and 99(1)(b), (c.1), (d), and (f) of the Act require border officers to form reasonable suspicion before strip searching travellers, examining imported or exported mail, and searching vehicles. Subsections 99.2(1) and 99.3(2) of the Act impose the same requirement to conduct searches of the person and intrusive examinations of goods in the custody or possession of people in customs controlled areas at Canadian international airports and marine ports.

[15] But subsection 99(1)(a) of the Act, whose constitutionality is at issue in this case and to which I refer in this decision as “the law,” does not require officers to meet an objective threshold before searching the contents of digital devices. This law provides, in relevant part, that “[a]n officer may, at any time up to the time of release, examine any goods that have been imported.” In other words, it authorizes Agency officers, without forming any reasonable suspicion that a border law has been violated, to search any goods that people bring across the border with them or otherwise import, until officers authorize those goods’ release from customs for use in Canada. The Agency relies on this law to search digital devices. In the past, until the Court of Appeal for Alberta decided Canfield in October 2020, most courts which had considered the question had treated the law as being constitutional: Canfield, at paras. 69, 163.

[16] Despite the law’s lack of an objective threshold, the Agency’s policy is that officers should only search digital devices if they can point to one or more objective facts that, taken together with other information, suggest that the device contains evidence of border law violations. The policy further provides that officers should also conduct “progressive” searches by targeting device-stored information relevant to their concerns about border law violations and intensifying their examination as their level of suspicion increases. Finally, the officers should take notes and record the areas and items searched. This policy guidance does not have the force of law.

(2) The Agency Searches Mr. Pike and Mr. Scott

[17] Agency officers relied on the law to search Mr. Pike’s and Mr. Scott’s digital devices when they returned to Canada in 2020. When the searches uncovered child pornography, the officers arrested them for the Customs Act offence of importing those illegal goods. The officers contacted the Peel Regional Police, who charged Mr. Pike and Mr. Scott with the Criminal Code offences of possessing and importing child pornography.

(a) Mr. Pike

[18] Jeremy Pike is a former teacher. Before the charges, he was already a registered sexual offender who had been sentenced to 14 years in prison and stripped of his teacher’s licence for sexually interfering with eight young boys and making and possessing child pornography depicting that abuse: R. v. J.P., 2008 ONCJ 484, aff’d, 2012 ONCA 549. After his release, he left Canada and spent seven months visiting several developing countries, some of which are known to be destinations for travellers intending to sexually exploit children.

[19] Mr. Pike returned to Canada alone in February 2020 with 14 digital devices and numerous children’s items in his luggage. Agency officers referred him for secondary examination. He looked nervous. After initially claiming that he was teaching adults, he ultimately told Agency Border Services Officer Taras Grama that he had been teaching seven-year-old boys in Indonesia. Officer Grama suspected Mr. Pike and decided to search all his devices. He demanded Mr. Pike’s computer password, which Mr. Pike provided, and searched all 11 working devices. He did not inform Mr. Pike of his right to counsel until 1 hour and 24 minutes into the search. Nearly 50 minutes later, he found child pornography on one of the devices, arrested Mr. Pike for violating the Customs Act, and contacted Peel Regional Police, who found over 1,600 images depicting the sexual abuse and exploitation of children on two of Mr. Pike’s devices.

(b) Mr. Scott

[20] David Scott is a 77-year-old retired business executive. He returned to Canada in June 2020 after spending six months at his vacation home in Belize. Agency Border Services Officer Jagdeep Sodhi approached him at the baggage claim and began asking him questions because he was travelling alone and had arrived via Houston. That airport is a hub for Caribbean and South America flights that, in the Agency’s view, pose a high risk for illegal drugs, child pornography, sexual exploitation, and national security issues.

[21] Officer Sodhi decided to conduct a secondary examination of Mr. Scott. He searched his luggage and found 13 digital devices. Mr. Scott was pacing, appeared nervous, and asked to repack the hard drives after Officer Sodhi encountered them. He said he had so many devices because he stored movies on them. Officer Sodhi thought this was implausible and suspected that he was importing child pornography. He demanded the password to Mr. Scott’s digital devices, which Mr. Scott provided. He and his fellow officers then searched his three cell phones, one of the two tablets, and some USB keys, but did not search any of the eight hard drives. They did not inform Mr. Scott that he was detained, of the reasons for his detention, or of his right to counsel until after they found a folder that contained written child pornography 55 minutes into the search. Nineteen minutes later, after finding child pornography images, Officer Sodhi arrested Mr. Scott for violating the Customs Act. The Agency contacted Peel Regional Police, who found approximately 3,000 text, image, and video files depicting the sexual abuse and exploitation of children on his digital devices.

(3) The Trial Judge’s Decisions

[22] Mr. Pike and Mr. Scott challenged the law’s constitutionality and sought to exclude the files of child pornography found on their digital devices, arguing that the Agency violated their Charter rights. The provincial Crown and the Public Prosecution Service of Canada, a party in the trial-level proceedings and on appeal in respect of the constitutional issue, defended the law’s constitutionality.[^1]

[23] The trial judge ruled that the law was unconstitutional: R. v. Pike, 2022 ONSC 2297, 413 C.C.C. (3d) 529. He also ruled that the Agency breached Mr. Pike’s and Mr. Scott’s s. 8 Charter right to be free from unreasonable searches and seizures, as well as breaching their s. 10(b) Charter right to counsel and Mr. Scott’s s. 10(a) Charter right to be informed promptly of the reasons for his detention. He excluded from evidence the child pornography files from Mr. Pike’s devices and acquitted him: R. v. Pike, 2022 ONSC 2298. However, he admitted the files from Mr. Scott’s devices (R. v. Scott, 2022 ONSC 4496, 515 C.R.R. (2d) 33), convicted him of importing child pornography (R. v. Scott, 2022 ONSC 5823), and sentenced him to 23 months’ house arrest (R. v. Scott, 2023 ONSC 3023).

[24] The Crown now appeals Mr. Pike’s acquittal and seeks leave to appeal the sentence the trial judge imposed on Mr. Scott, and Mr. Scott appeals his conviction.

C. analysis

[25] These appeals raise four questions, namely whether:

1. The Customs Act’s authorization of searches of digital devices violates the Charter;

2. The Agency violated Mr. Pike’s and Mr. Scott’s right to counsel;

3. The evidence found on their digital devices is admissible; and

4. The trial judge erred by imposing a conditional sentence on Mr. Scott.

[26] As I will explain, the answer to each of these questions is yes.

(1) The Law Violates Travellers’ Right Against Unreasonable Searches

[27] Mr. Pike and Mr. Scott claim, and the trial judge ruled, that the Agency violated their s. 8 Charter right to be secure against unreasonable search and seizure. I agree. Like other travellers, Mr. Pike and Mr. Scott reasonably expect privacy in regard to their digital devices. The state’s invasion of that strong privacy interest was authorized by a law, but that law unreasonably offends s. 8 of the Charter. It does not strike a fair balance between the privacy interests of travellers and the state’s interests because it attaches the lowest possible threshold to one of the most intrusive searches imaginable. Because the Crown has not justified this limit under s. 1 of the Charter, the Agency violated Mr. Pike’s and Mr. Scott’s s. 8 rights when it searched their digital devices.

[28] At the outset, I commend the trial judge for his systematic and well-reasoned ruling on the constitutional challenge, with which I largely agree and which I would uphold. Canfield, which is the first Canadian appellate decision to declare the law unconstitutional, also served as a helpful guide. However, as I will explain, I part company with Canfield on the appropriate threshold for searches of digital devices. In my respectful view, a reasonable suspicion test, rather than the less demanding tailored search approach suggested in Canfield, is needed to properly protect privacy.

(a) The Legal Framework for Unreasonable Search Claims

[29] Section 8 of the Charter protects everyone’s “right to be secure against unreasonable search or seizure.” Its main purpose is to protect privacy, meaning the reasonable right that each of us enjoy to be left alone by the state. Privacy protections preserve our liberty and prevent government from prying into our lives. We have privacy interests not merely in our homes and other places where we live and work, but also in our bodies and in personal information about ourselves: R. v. Tessling, 2004 SCC 67, [2004] 3 S.C.R. 432, at paras. 12‑16, 20-24.

[30] A two-step test governs s. 8 claims. At step one, claimants must prove that the state engaged in a search or seizure. To prove a search, claimants must show that the state invaded their reasonable expectation of privacy. Courts determine this expectation’s existence by considering the subject matter the state is targeting, the claimants’ interest in that subject matter, their subjective expectation of privacy, and the objective reasonableness of that subjective expectation: R. v. Bykovets, 2024 SCC 6, 489 D.L.R. (4th) 1, at paras. 29-31. If claimants show the state searched them without obtaining a judicial warrant authorizing the search, then the search is presumptively unreasonable. At step two, the state must rebut this presumption and prove on a balance of probabilities that the search was reasonable by showing that a reasonable law authorized it and that the state carried it out reasonably. If the state does not meet this burden, then the search limits s. 8 rights and the state must justify that limit under section 1 of the Charter. If the state fails to do so, then the search is unconstitutional: Goodwin v. British Columbia (Superintendent of Motor Vehicles), 2015 SCC 46, [2015] 3 S.C.R. 250, at paras. 48, 56, 79.

[31] This court reviews the trial judge’s legal conclusions and findings of mixed fact and law concerning these constitutional issues for correctness. We owe deference only to purely factual findings that can be isolated from the constitutional analysis, such as the trial judge’s findings concerning border officers’ interactions with Mr. Pike and Mr. Scott: Société des casinos du Québec inc. v. Association des cadres de la Société des casinos du Québec, 2024 SCC 13, 491 D.L.R. (4th) 385, at para. 45, per Jamal J., and at paras. 94-97, per Côté J. (concurring).

(b) Travellers Reasonably Expect Privacy in Their Digital Devices

[32] The trial judge ruled, and the Crown concedes, that step one of the test is met and that the Agency’s examination of Mr. Pike’s and Mr. Scott’s digital devices was a search under s. 8 of the Charter because travellers have a reasonable expectation of privacy in the contents of those devices. I agree. Travellers have a direct interest and subjective expectation of privacy in the contents of their digital devices, which are the subject matter of the examination by the state agents: R. v. Cole, 2012 SCC 53, [2012] 3 S.C.R. 34, at paras. 41-43. That expectation of privacy is objectively reasonable because digital devices “are a repository for an almost limitless universe of information,” including sensitive and confidential information concerning travellers’ lifestyle and personal choices: Bykovets, at paras. 54-55; see also Cole, at paras. 44-48. Travellers’ exercise of control and ownership over their digital devices strengthens this reasonable expectation (R. v. Law, 2002 SCC 10, [2002] 1 S.C.R. 227, at para. 16), which the border context tempers, but does not eliminate: R. v. Monney, 1999 CanLII 678 (SCC), [1999] 1 S.C.R. 652, at paras. 35-36, 42 (“Monney (SCC)”).

(c) The Law Authorizes Unreasonable Searches

[33] Because the Agency did not obtain a warrant to search Mr. Pike’s and Mr. Scott’s digital devices, the Crown has the burden to prove that a reasonable law authorized those searches. I agree with the trial judge that it has not met this burden and that the Agency accordingly limited Mr. Pike’s and Mr. Scott’s s. 8 Charter rights. The Crown is correct that the law, s. 99(1)(a) of the Customs Act, authorizes the searches. But this law is unreasonable. It adopts the lowest possible threshold, namely subjective good faith purpose, rather than the reasonable suspicion threshold that the Charter requires. I next explain which searches the law authorizes, why the law’s subjective good faith purpose threshold is unreasonable, and why reasonable suspicion is the correct threshold.

(i) The Law Authorizes Low Threshold Digital Device Searches

The Law Authorizes the Searches

[34] The state cannot search people unless statutes or judge-made rules authorize it to do so, and it complies with those laws’ requirements and restrictions. This rule limits state power to protect individual freedom and the rule of law: R. v. Caslake, 1998 CanLII 838 (SCC), [1998] 1 S.C.R. 51, at para. 12; R. v. Lee, 2017 ONCA 654, 351 C.C.C. (3d) 187, at para. 110, per Pardu J.A. (concurring).

[35] The law at issue, s. 99(1)(a) of the Customs Act, authorizes border officers to search any goods that travellers bring across the border with them until officers authorize those goods’ release from customs for use in Canada. The Crown argues that the law authorizes digital device searches because the electronic files those devices contain are goods. The intervener, the Criminal Lawyers’ Association (Ontario) (the “CLA”), disagrees and argues that the law only authorizes searches of physical goods. I agree with the Crown. The law’s text, context, and purpose establish that it authorizes border officers to search the contents of those digital devices, which are electronic documents: Canada v. Loblaw Financial Holdings Inc., 2021 SCC 51, 464 D.L.R. (4th) 244, at para. 41.

[36] First, the text of the Customs Act establishes that electronic files are goods. Section 2 provides that documents are goods. At the relevant time, when Parliament enacted the Customs Act in 1985 (Perka v. The Queen, 1984 CanLII 23 (SCC), [1984] 2 S.C.R. 232, at pp. 264-265), “document” meant something that records information, no matter the recording medium: The Oxford English Dictionary, 2nd ed. (Oxford: Clarendon Press, 1989), vol. IV, at p. 916; Mouammar v. Brunner (1978), 1978 CanLII 1676 (ON SC), 19 O.R. (2d) 59 (S.C.), at pp 60-61; Hill v. R., [1945] 1 K.B. 329, at pp. 332-333. This covers electronic files because they record information on an electronic medium, as explained in Prism Hospital Software Inc. v. Hospital Medical Records Institute (1991), 1991 CanLII 11770 (BC SC), 62 B.C.L.R. (2d) 393 (S.C.), at paras. 6-12. Parliament made clear that it intended this broad meaning by defining goods to include not just some documents in certain forms, but “any document in any form” (Customs Act, s. 2, (emphasis added)), meaning “no matter which” form: National Bank of Greece (Canada) v. Katsikonouris, 1990 CanLII 92 (SCC), [1990] 2 S.C.R. 1029, at p. 1038. This broad, clear, and precise ordinary meaning shows that “document” includes new and emerging forms of recording information, such as electronic files, and plays a dominant interpretive role: British Columbia Telephone Co. v. M.N.R. (1992), 139 N.R. 211 (Fed. C.A.), at paras. 9 n.5, 15‑17, citing Perka, at p. 265; Loblaw, at para. 41.

[37] The definition of “document” in the French version of the Customs Act confirms that electronic documents are goods. Per the French version, a document is anything that records data that can be read or understood by a person, a computer system, or another device.[^2] This unambiguous definition expressly includes electronic documents, resolves any potential ambiguity in the English version, and confirms that the common meaning of document includes electronic documents: R. v. Mac, 2002 SCC 24, [2002] 1 S.C.R. 856, at paras. 5-6. It also provides the distinct, specific authorization to search electronic documents that the CLA submits is necessary under the Supreme Court’s decision in R. v. Vu, 2013 SCC 60, [2013] 3 S.C.R. 657. The CLA’s argument that the French definition does not govern because it is broader than the undefined English term “document” must fail. Even assuming for the sake of argument that “document” has a narrower potential meaning in English as well as the broader meaning I outlined above, this would only establish that the English version is ambiguous. The French version would still govern in this scenario because it is unambiguous. The narrower official language version only governs if, unlike here, both versions are ambiguous or two unambiguous versions clash: R. v. Daoust, 2004 SCC 6, [2004] 1 S.C.R. 217, at paras. 28-29.

[38] Second, the statutory context confirms that electronic documents are goods. The Customs Act should be interpreted harmoniously with the Customs Tariff, S.C. 1997, c. 36, because they form an integrated border regulatory scheme by which the Act grants border officers authority to prevent the entry of goods that the Tariff prohibits and to ensure that travellers pay duties on other goods at Tariff rates: Bell ExpressVu Limited Partnership v. Rex, 2002 SCC 42, [2002] 2 S.C.R. 559, at para. 27. Tariff Schedule Item 9899.00.00 lists the electronic form of child pornography as a good. Because the Tariff adopts the Customs Act’s definition of goods, this shows that Parliament accepted that electronic documents are goods under that Act: Tariff, s. 4; R. v. Gibson, 2017 BCPC 237, at paras. 44, 90.

[39] The CLA’s context argument that the limited class rule and associated words rule restrict searches to physical documents therefore must fail. Per the limited class rule, general terms refer to the same class of items as specific terms that precede them: National Bank of Greece, at p. 1040. But this rule does not apply to s. 2’s definition of “goods” because “document” is a specific term like the references to “conveyances” and “animals” that precede it, not a general term: British Columbia Telephone, at para. 21. The CLA next turns to the associated words rule, which provides that the meaning of terms connected by “and” or “or” that play a similar role can be narrowed to their widest common denominator: Opitz v. Wrzesnewskyj, 2012 SCC 55, [2012] 3 S.C.R. 76, at para. 41. The CLA submits this rule applies because, after authorizing officers to examine imported goods, s. 99(1)(a) also creates another search power that authorizes officers to open packages or containers of imported goods and take samples of them. But this rule does not apply either because the two distinct search powers that s. 99(1)(a) creates lack a meaningful common denominator: Daoust, at paras. 60‑61. That is because the goods that the law authorizes officers to examine need not be stored in containers that can be sampled from.

[40] Third, statutory purpose favours treating electronic documents as goods. The purpose of the Act is to regulate the cross-border movement of people and goods: Martineau, at para. 25. This supports interpreting “goods” broadly to cover all items whose cross-border movement the state has an interest in regulating: Monney (SCC), at paras. 26-28. So does the related functional equivalence principle, which favours applying legislation to new technologies that play a similar role to the technologies in existence when it was passed: Lumberland Inc. v. Nineteen Hundred Tower Ltd., 1975 CanLII 196 (SCC), [1977] 1 S.C.R. 581, at p. 593. Courts have long used this principle to interpret the term “document” to cover new technologies like contemporary electronic documents that, similar to the written paper documents which were widespread when Parliament passed the Act in 1985, record information: Fox v. Sleeman (1897), 17 P.R. 492 (Div. Ct.) (photographs are documents); Mouammar, at pp. 60-61 (tapes are documents); Prism, at paras. 6‑12 (electronic files are documents); R. v. Whittaker, 2010 NBPC 32, 367 N.B.R. (2d) 334, at para. 8 (same). In contrast, restricting goods to physical items undermines the Act’s purpose by denying officers the power to search for electronic child pornography under the Act’s many powers to search goods.

[41] Fourth, the CLA’s reliance on the interpretive presumptions of Charter conformity and against extra-territoriality is misplaced. The former has no role to play because the common meaning of the French and English versions of the Act is unambiguous: Bell ExpressVu, at paras. 28, 62, 66. Neither does the latter because the law only authorizes border officers to search goods that travellers bring with them into Canada: R. v. Desautel, 2021 SCC 17, 456 D.L.R. (4th) 1, at para. 43. The CLA’s contrary claim that the law authorizes border officers to search data stored in other countries that is accessible from, but has not been downloaded to, travellers’ digital devices is wrong. The law does not authorize border officers to do so because travellers are not importing that data into Canada when they cross the border: Gibson, at paras. 91-92. Clearer language would be required, such as s. 487(2.1) of the Criminal Code’s authorization to search data “available to” a computer system, to enable searches of data not stored on but accessible from a digital device.[^3]

The Law Sets a Low Threshold

[42] Because the law authorizes searches of digital devices, we must determine which limits, if any, it imposes on that search power. The Crown argues that the law guards against broad invasions of privacy by implicitly requiring the search to be proportional in its scope to a valid border law enforcement purpose and that it is to be conducted manually. This argument fails. While the law does require border officers to act in good faith and pursue a valid border law purpose, it does not require searches to be proportional to that purpose or to be conducted manually.

[43] Border officers must use the law’s search power in good faith and for a valid border law enforcement purpose. The caselaw recognizes this implied requirement, which the law shares with other search powers: R. v. Singh, 2014 ONSC 5658, 317 C.C.C. (3d) 446, at paras. 49-51; Comité paritaire de l’industrie de la chemise v. Potash, 1994 CanLII 92 (SCC), [1994] 2 S.C.R. 406, at p. 425. It is a particular application of the bedrock rule of law principle from Roncarelli v. Duplessis that legislative grants of power to public officials “necessarily impl[y] good faith in discharging public duty” consistent with the “perspective within which a statute is intended to operate”: 1959 CanLII 50 (SCC), [1959] S.C.R. 121, at p. 140, per Rand J. Here, that perspective is the Customs Act’s purpose of regulating people and goods’ cross-border movements: Martineau, at para. 25.

[44] This good faith purpose requirement imposes modest but real limits. As with other search powers, it imposes a legitimate target rule that requires officers to look for evidence of border law violations: R. v. Jones, 2011 ONCA 632, 107 O.R. (3d) 241, at paras. 42-43 (“Jones (2011)”); R. v. Appleton, 2011 CarswellOnt 11191 (S.C.), at para. 12. This bars officers from indiscriminately scouring the devices and examining programs or files for no legitimate border-related reason: Vu, at para. 61. Border officers also cannot make search decisions motivated or influenced by racial profiling: R. v. Dudhi, 2019 ONCA 665, 147 O.R. (3d) 546, at para. 64; R. v. Smith (2004), 2004 CanLII 46666 (ON SC), 26 C.R. (6th) 375 (Ont. S.C.), at paras. 29-30, 34. Nor can they use the law as a “mere ruse or pretext” to investigate non-Customs Act crimes: R. v. Mayor, 2019 ONCA 578, 378 C.C.C. (3d) 453, at paras. 7-9; Appleton, at para. 12.[^4]

[45] That is as far as the good faith purpose requirement goes. It cannot be stretched beyond its modest rule-of-law-preserving role of ensuring that border officers act in good faith and consistent with the Customs Act’s purpose to impose a proportionality rule or other prohibitions on more intrusive searches. To the contrary, as the trial judge found, the good faith purpose requirement permits “wide-ranging inspection[s]” of digital files if necessary (Jones (2011), at para. 43) because the information officers seek can sometimes “be[] found almost anywhere in the [device]” (Vu, at para. 60). As the trial judge also found, Officer Sodhi’s searches of Mr. Scott’s tablet demonstrate this. He testified that he examined the Desktop, Documents, Downloads, and Pictures folders, which are not limited to recent communications, but contain a broad trove of highly sensitive personal information, because he expected travellers would store files, including child pornography, there.

[46] Parliament could have, but did not, impose more limits. We know that because it did so elsewhere in the Act by adding a customs controlled area search power that, unlike s. 99(1)(a), only authorizes non-intrusive suspicion-less goods searches: An Act to amend the Customs Act and to make related amendments to other Acts, S.C. 2001, c. 25, s. 60; Customs Act, s. 99.3(1). Because we are judges and not legislators, we cannot rewrite the Act by importing that limit into s. 99(1)(a): TELUS Communications Inc. v. Wellman, 2019 SCC 19, [2019] 2 S.C.R. 144, at paras. 79-80.

[47] That is why the Crown’s argument that we should read into the law the court-created cell phone search restrictions from R. v. Fearon, 2014 SCC 77, [2014] 3 S.C.R. 621, fails. Fearon modified judge-made rules concerning cell phone searches incident to arrest: at para. 58. It is an example of courts exercising their limited power to incrementally adapt judge-made police powers where legislation is absent: Fleming v. Ontario, 2019 SCC 45, [2019] 3 S.C.R. 519, at para. 42; see Fearon, at para. 84. Fearon does not apply here because, by enacting the law, Parliament has entered the field. This shifts our judicial role from adapting judge-made rules to interpreting Parliament’s rules and reviewing them for constitutional compliance.

[48] The Crown’s argument that the law prevents forensic searches and only authorizes manual searches also fails. This argument wrongly assumes that the law incorporates the Agency’s practice of only conducting manual searches under it. But nothing in the law’s text incorporates this restriction. The Agency’s practice of conducting manual searches does not transform this restriction into a binding legal rule, and we cannot rewrite the law by reading it in: R. v. Beaudry, 2007 SCC 5, [2007] 1 S.C.R. 190, at paras. 45-46.

(ii) The Law’s Low Search Threshold Is Unreasonable

The Reasonableness Test

[49] Laws authorizing privacy intrusions are reasonable if they strike a proper balance between people’s privacy interests and the state’s interests. Courts assess search laws' reasonableness by considering their: (1) intrusiveness, (2) reliability, (3) oversight mechanisms, (4) purpose, and (5) criminal or regulatory nature: Goodwin, at paras. 55-57, 67-68.

[50] The Supreme Court gave specific guidance on border search laws’ reasonableness in Simmons. Simmons held that border searches generally require lower thresholds than non-border searches because of the state’s strong border regulatory interests, consequent need for search powers, and travellers’ reduced expectations of privacy: at pp. 527-529.

[51] Simmons grouped border searches into three categories. These are, in ascending order of intrusiveness: (1) routine questioning, searching luggage and belongings, and frisk or pat down searches; (2) strip searches; and (3) body cavity searches: at pp. 516-517.

[52] Simmons and subsequent cases also attached thresholds to these categories by applying the principle that more intrusive searches require greater protection and, thus, higher thresholds: Simmons, at p. 517. The lowest and the only purely subjective standard, good faith purpose, applies to category one routine questioning and luggage, frisk, and pat down searches: Singh, at paras. 49-51.[^5] Because this standard does not impose an objective threshold, officers can perform these searches on all travellers if they do so for a legitimate border-related reason and neither make search decisions motivated or influenced by racial profiling nor use these searches as a mere ruse or pretext to investigate non-Customs Act crimes: Appleton, at para. 12; Smith, at paras. 29-30, 34. As Mr. Scott and the CLA submit, these searches merit reduced protection not because the Agency regularly conducts them, but because their subject matter engages reduced privacy expectations.

[53] Next, Simmons required that officers meet an intermediate standard, reasonable suspicion, before conducting strip searches: Simmons, at pp. 517, 529.[^6] Unlike good faith purpose, reasonable suspicion must be based on objective facts that show a reasonable possibility of a border law violation: R. v. Chehil, 2013 SCC 49, [2013] 3 S.C.R. 220, at para. 27.

[54] Finally, Simmons held that body cavity searches require an even greater but unspecified degree of protection: at p. 517; see also Monney (SCC), at para. 39. I would suggest that these searches require at least reasonable and probable grounds because that is the next threshold after reasonable suspicion in ascending order of protectiveness: Chehil, at para. 24. Reasonable and probable grounds also require objective facts but, unlike reasonable suspicion, those facts must show a reasonable probability of crime rather than a mere possibility: Chehil, at para. 27.

[55] Courts determine which threshold is reasonable for a border search by mapping that search onto a Simmons category. Because each category is discrete and not a continuum, each includes searches of varying intrusiveness. For instance, pocket and frisk searches both qualify as category one even though, as explained in R. v. Mann, 2004 SCC 52, [2004] 3 S.C.R. 59, at para. 49, the former is more intrusive than the latter. Similarly, passive bedpan vigils qualify as category two searches even though, unlike strip searches, they intrude on travellers’ strong privacy interest in their bodily functions: R. v. Hudson (2005), 2005 CanLII 47233 (ON CA), 77 O.R. (3d) 561 (C.A.), at paras. 26-30; Monney (SCC), at paras. 37-48.

[56] But Simmons does not address what category or threshold searches of digital devices require. It never mentions them. This is not surprising because the Supreme Court decided Simmons in the 1980s, when personal digital devices were in their infancy, were rarely transported across borders, and did not contain anywhere near the immense amounts of information they do today: Canfield, at paras. 28, 37. I disagree with Canfield that deciding what protections the Charter grants in relation to digital devices at the border requires revisiting Simmons under the exceptions to the doctrine of legal precedent set out in Canada (Attorney General) v. Bedford, 2013 SCC 72, [2013] 3 S.C.R. 1101, at para. 42: Canfield, at paras. 25-38. Rather, resolving this question does not trigger that doctrine because, as the trial judge reasoned, Simmons never decided that issue: see R. v. Kirkpatrick, 2022 SCC 33, 471 D.L.R. (4th) 440, at para. 85.

[57] Because Simmons does not answer this question, I apply the five-factor Goodwin reasonableness test to determine whether the law is reasonable. I consider the Simmons categories and border search guidance when applying this test because they are relevant to several Goodwin factors.

The Law Is Unreasonable

[58] I agree with the trial judge that the law is unreasonable because its purely subjective good faith purpose test undervalues travellers’ privacy interests in the contents of their digital devices. That low threshold, coupled with the law’s unreliability and lack of oversight mechanisms, authorizes the Agency to invade the privacy of numerous innocent travellers who have little recourse to challenge its actions. The law’s border control purposes and regulatory dimension temper the need for a high threshold but do not make Parliament’s choice of the lowest possible threshold reasonable.

[59] First, beginning with the initial Goodwin factor, being the law’s intrusiveness, searches of digital devices require a meaningful threshold because they target highly private information that s. 8 strongly protects. In a 15-year line of cases running from R. v. Morelli, 2010 SCC 8, [2010] 1 S.C.R. 253, through this year’s Bykovets decision, the Supreme Court has held that people have “unique and heightened privacy interests in personal computer data” because it can “expose deeply revealing information”: Bykovets, at para. 55. In fact, it is hard to imagine a more intrusive search because, as our “constant companion[s],” digital devices record immense quantities of information concerning our friendships, love and sexual desires, finances, medical information, interests, and political and social views, as well as our future plans: Bykovets, at paras. 1, 54; see Cole, at para. 47. They contain “a digital record of nearly every aspect of [our lives” and “could just as easily be called cameras, video players, rolodexes, calendars, tape recorders, libraries, diaries, albums, televisions, maps, or newspapers”: Riley v. California, 573 U.S. 373 (2014), at pp. 393-395. Not only that, but they generate information without users’ knowledge, retain data that users try to destroy, and let users access information stored on remote servers: Vu, at paras. 42-44. While travellers’ reduced expectation of privacy at the border tempers the strength of these privacy interests, they remain strong as the trial judge recognized.

[60] In addition to these strong informational privacy interests, digital device searches engage travellers’ personal privacy with respect to their bodies. This form of privacy is not limited to physical touching but includes visual access to the body, such as by viewing photos or videos of the naked or undressed body: R. v. Jarvis, 2019 SCC 10, [2019] 1 S.C.R. 488, at paras. 62, 65, 67 (“Jarvis (2019)”). As a result, digital device searches risk engaging personal privacy because people often store intimate images on their devices: R. v. J.R., 2018 ONCJ 851, at para. 92. I cannot overlook this risk because it has already materialized in a reported case in which border officers encountered and, albeit without improper intent, scrolled through intimate images while searching a device for contraband: Gibson, at paras. 29, 189. This risk heightens the law’s intrusiveness because viewing these images invades basic human dignity (R. v. Walsh, 2021 ONCA 43, 154 O.R. (3d) 263, at para. 67; see also Jarvis (2019), at para. 65), even where Agency officers scroll through them casually and without improper intent (Justin Doll, “Searches of the Person: A New Approach to Electronic Device Searches at Canadian Customs” (2020) 18 Can. J. L. & Tech. 91, at p. 112).

[61] I agree with the trial judge that the law’s intrusiveness shows that it needs a higher category and threshold than the subjective good faith purpose test it adopts, the same test that governs Simmons category one luggage searches. That test is insufficient for digital device searches, which the Supreme Court and this court have held are fundamentally more intrusive than luggage searches because, unlike luggage, digital devices contain “an almost unlimited universe of information”: Vu, at para. 41, quoting R. v. Mohamad (2004), 2004 CanLII 9378 (ON CA), 69 O.R. (3d) 481 (C.A.), at para. 43. This holding reflects reality. While most people cannot and do not carry all their physical mail, pictures, books, and papers in a suitcase, digital devices routinely store all that information and more: Riley, at pp. 393-395. That is why digital devices cannot be lumped in with luggage searches. Rather, as Simmons holds, “it is obvious that the greater the intrusion, the greater must be the justification and … constitutional protection”: at p. 517.

[62] The Crown’s counterargument that internal limits temper the law’s intrusiveness collapses because, as I explained, its principal premise, being that the law imposes proportionality and manual search limits, is incorrect. None of the law’s actual limits save it. The good faith purpose limit merely prevents officers from abusing their power for ulterior purposes, but it does not bar them, for instance, from conducting highly intrusive searches of travellers whom they have no basis to suspect in the hopes that they might find contraband. Likewise, as Mr. Pike submits, the limit preventing border officers from searching documents accessible from, but not downloaded to, the device does not prevent officers from searching the extensive amounts of highly sensitive personal information that are downloaded to those devices, such as the contents of the Desktop, Documents, Downloads, and Pictures folders that Officer Sodhi searched: Vu, at para. 41.

[63] The Crown’s related counterargument that travellers’ “choice” to travel with digital devices merits a lower threshold also fails. Because digital devices are our “constant companion[s]” (Bykovets, at para. 1), travellers need to bring them across borders to work and communicate. As the trial judge ruled, leaving them behind is not a meaningful choice. Neither is declining to leave and re-enter Canada, which, as the intervener Canadian Civil Liberties Association (the “CCLA”) submits, is not merely a choice but a section 6 Charter right. Just as “Canadians are not required to become digital recluses” to preserve their privacy (R. v. Jones, 2017 SCC 60, [2017] 2 S.C.R. 696, at para. 45), they also should not have to surrender the ability to enter and leave Canada with an indispensable instrument of modern life.

[64] Second, the next two Goodwin factors, reliability and oversight mechanisms, together with the risk of racial profiling, confirm the case for the unreasonableness of the law and the need for a higher threshold. Beginning with reliability, the searches the law authorizes mostly miss the mark due to its low threshold. Approximately 62% of the 31,579 digital device searches the Agency conducted between 2017 and 2020 found no evidence of any border law violations. This significant invasion of the privacy of many innocent travellers is a strong sign that the law is unreasonable: Goodwin, at para. 67.

[65] Linked to these reliability concerns is the risk that border officers will, consciously or unconsciously, target travellers based on their race or racial stereotypes. We must be cognizant of this risk because it is still an everyday reality in the lives of many racialized people in our diverse province and country: R. v. Le, 2019 SCC 34, [2019] 2 S.C.R. 692, at para. 80. This is especially true since Agency officers targeted Mr. Pike and Mr. Scott because they arrived from countries that the Agency associates with drug trafficking and child sexual exploitation. These country-specific concerns create a risk of racially profiling travellers who are nationals of, or have family connections to, those countries or simply share the same ethnicity or skin colour as those countries’ inhabitants.

[66] The law’s low threshold increases this risk because, as the CCLA submits, low threshold powers are the easiest for officers to wield to target, whether intentionally or not, racialized and disadvantaged people: R. v. Landry, 1986 CanLII 48 (SCC), [1986] 1 S.C.R. 145, at p. 186, per La Forest J. (dissenting), rev’d, R. v. Feeney, 1997 CanLII 342 (SCC), [1997] 2 S.C.R. 13; see also Le, at para. 87. While the law’s good faith purpose test offers some protection against this risk, its subjective nature makes that risk harder to detect because officers do not have to point to objective facts to justify the search and help negate the possibility of discrimination: Chehil, at paras. 3, 25, 30; see also R. v. Ladouceur, 1990 CanLII 108 (SCC), [1990] 1 S.C.R. 1257, at p. 1267, per Sopinka J. (concurring).

[67] The third factor, oversight, points in the same direction. After-the-fact oversight is especially important because, unlike in Simmons where the strip search law empowered travellers to seek judicial or administrative review before the officer conducted the search, this law lacks any prior authorization mechanism: Goodwin, at para. 71; see Simmons, at p. 529. But the law’s low threshold hinders after-the-fact judicial oversight where searches lead to charges because, unlike the reasonable suspicion standard in Chehil, it does not allow courts to ensure that officers had an objective basis to search: Goodwin, at para. 70, citing Chehil, at para. 25. Instead, all courts can do is check that officers were, in their own minds, acting for a good faith purpose.

[68] The oversight problem is even worse for searches that do not lead to charges. The Customs Act statutory review mechanisms are inapplicable to digital device searches that do not lead to either seizures of goods or to an Agency determination that travellers are importing prohibited goods—in other words, to searches of travellers who are innocent of any wrongdoing: Customs Act, ss. 126.1-137. The other review mechanisms do not lead to meaningful remedies. Complaints under the Privacy Act, R.S.C. 1985, c. P-21, merely authorize the Privacy Commissioner to make recommendations that the Agency is free to disregard (see s. 35(1)), and the Agency’s online complaints mechanism is an internal accountability tool with no legal consequences. The law thus creates the very problem of difficult-to-review discretionary searches that Goodwin warns about: at para. 70.

[69] Third, the last two Goodwin factors, the law’s purpose and nature, do not justify its low threshold. Instead, they merely temper the strong privacy protections the Charter grants digital devices outside the border context.

[70] The law’s important border purpose does not excuse its purely subjective threshold. Per Simmons, border search powers advance state interests in preventing the entry of unauthorized persons and prohibited goods, and in ensuring travellers pay tariffs on authorized goods: at pp. 527-528. Border regulation is vital to Canada’s sovereignty and helps ensure national security, public health, tax collection, and crime control, among other interests: Jones (2006). These interests, along with people’s reduced privacy expectations when traveling internationally, temper privacy protections at the border: Simmons, at pp. 527-529.[^7]

[71] However, I agree with the trial judge that these factors do not justify lumping searches of digital devices into Simmons category one along with the very briefcases the Supreme Court and this court have held are fundamentally different from those digital devices. That would undercut Simmons’ holding that highly intrusive searches—formerly, strip and body cavity searches, and now digital device searches—deserve greater protection and cannot be lumped into category one: at pp. 516-517. While category one is broad, the differences it accommodates are less fundamental, such as an inner pocket search’s greater intrusiveness relative to an outer clothing pat down: Hudson, at paras. 26-30; see Mann, at para. 49.

[72] The Crown’s counterargument, that the law’s border purposes and context justify its low threshold, overshoots these factors’ measured effect. They do not transform the border into a Charter-free zone: Nagle, at para. 81. Nor do they reduce all border search tests to the lowest threshold, being good faith purpose. Instead, they temper privacy protections by moderating the threshold, such as by lowering the strip search test from the usual non-border threshold, reasonable and probable grounds, to reasonable suspicion at the border: Simmons, at pp. 528-529; see R. v. Golden, 2001 SCC 83, [2001] 3 S.C.R. 679, at para. 105. The same is true of digital device searches. Outside the border, the state generally needs a warrant backed by reasonable and probable grounds to search digital devices: Vu, at para. 48. While border purposes and context moderate that standard, allowing these factors to reduce it to the lowest possible standard is a bridge too far.

[73] The state’s less strong interest in searching for electronic documents compared to physical contraband confirms this conclusion. Simmons held that the state’s interest in suppressing the trafficking of drugs produced in other countries that had to be transported across the physical border to enter Canada justified the strip search law: at pp. 526-529. In contrast, digital contraband, even when downloaded to a device, is usually also stored on external servers and can be electronically transmitted into Canada, a mode of transmission that the Agency admits it has no mandate to control. I agree with the CCLA that this meaningfully weakens the strength of the state’s interests, as explained in United States v. Smith, 673 F. Supp. 3d 381 (S.D.N.Y. 2023), at pp. 394-396, by Professor Steven Penney in his article “‘Mere Evidence’? Why Customs Searches of Digital Devices Violate Section 8 of the Charter” (2016) 49 U.B.C. L. Rev. 485, at p. 510, and by Professor Robert Diab in his article “Protecting the Right to Privacy in Digital Devices: Reasonable Search on Arrest and at the Border” (2018) 69 U.N.B.L.J. 96, at p. 122.[^8]

[74] The law’s regulatory dimension does not save it either. Regulatory schemes often need search powers to function: Comité paritaire, at p. 422. Those powers may attract lower standards than criminal laws because they target conduct that is not inherently wrongful but requires regulation to protect the public: Goodwin, at para. 60; R. v. Wholesale Travel Group Inc., 1991 CanLII 39 (SCC), [1991] 3 S.C.R. 154, at pp. 218-219, per Cory J. That is true to a degree here because the law advances regulatory interests in collecting border taxes and preventing unauthorized people and prohibited goods from entering: Simmons, at pp. 514-517, 528-529.

[75] But the Supreme Court has stated the obvious—border search powers also have a criminal dimension. As Justice Bertha Wilson explained in R. v. McKinlay Transport Ltd., 1990 CanLII 137 (SCC), [1990] 1 S.C.R. 627, at pp. 647-648, Simmons “address[es] … searches or seizures in a criminal or quasi-criminal context,” not merely regulatory searches. That is because, as well as advancing border regulatory interests, border searches also target inherently wrongful conduct like importing child pornography or illegal drugs. As Mr. Pike submits, these are serious crimes under the Criminal Code and Controlled Drugs and Substances Act. They can result in significant jail sentences even when prosecuted under the Customs Act that the Agency enforces, as described in R. v. Basov, 2015 MBCA 22, 315 Man. R. (2d) 222. For these reasons, the state is “the singular antagonist against the individual” when it conducts border searches as well as a regulator of cross-border travel to protect the public: R. v. Monney (1997), 1997 CanLII 979 (ON CA), 153 D.L.R. (4th) 617 (Ont. C.A.), at para. 134, per Weiler J.A. (dissenting), rev’d, Monney (SCC). This factor favours requiring stronger protections: Jarvis (2002), at paras. 59, 62. I thus reject the Crown’s argument that the Agency’s border officers operate in a purely regulatory role that is separate from its criminal branch or policing. They do not because, in addition to their regulatory responsibilities, border officers search for evidence of and arrest people for committing the Customs Act offence of importing child pornography into Canada, as they did here before contacting the Peel Regional Police.

[76] I also agree with the trial judge and the CCLA that the law’s intrusiveness blunts the moderating impact of its regulatory dimension. Unlike regulatory laws that target less private information like business records, the law targets information that, like personal papers, is highly private: British Columbia Securities Commission v. Branch, 1995 CanLII 142 (SCC), [1995] 2 S.C.R. 3, at para. 62. This favours a higher standard even in a regulatory context: Baron v. Canada, 1993 CanLII 154 (SCC), [1993] 1 S.C.R. 416, at pp. 443-445; Jarvis (2002), at paras. 61-62.

[77] For all these reasons, I agree with the trial judge and Canfield that the law strikes the wrong balance between the state’s interests and travellers’ privacy interests and, accordingly, that the Agency limited Mr. Pike’s and Mr. Scott’s s. 8 rights. The law applies the lowest possible border search threshold to the very searches of digital devices that the Supreme Court has characterized as some of the most intrusive searches imaginable. The important purposes and regulatory goals the law serves temper travellers’ privacy interests but do not justify eviscerating them. Because the border is not a Charter-free zone, it is also not an almost-anything-goes zone for highly intrusive searches like digital device examinations.

(iii) The Law Needs a Reasonable Suspicion Threshold

[78] For all the same reasons, the Charter requires a reasonable suspicion threshold to protect the strong privacy interests of travellers in their digital devices, as enunciated in category two of Simmons. I agree with the trial judge that this threshold is the constitutional minimum.

Why Defining the Threshold Matters

[79] It is our responsibility as judges to define this threshold. Canfield declined to do so, reasoning that it is complex and should be left to Parliament and subsequent cases: at para. 75. But in the seminal s. 8 case, Hunter v. Southam, the Supreme Court told Parliament what search threshold would be Charter-compliant when it struck down legislation that imposed a lower threshold: 1984 CanLII 33 (SCC), [1984] 2 S.C.R. 145, at p. 168. This makes sense because, when judges exercise the momentous power to declare laws to be of no force and effect because they are inconsistent with the Charter, they have a corresponding responsibility, where possible, to outline clear standards that legislatures can meet to accomplish their policy objectives in a constitutionally compliant manner. Even if legislatures disagree with those standards, defining them enables legislatures to make an informed response, consistent with the principles discussed in Vriend v. Alberta, 1998 CanLII 816 (SCC), [1998] 1 S.C.R. 493, at paras. 136-139. In my respectful view, it would not be helpful to tell Parliament that, while the threshold it set is too low, we are not prepared to say which higher threshold would pass constitutional muster.

Reasonable Suspicion is the Correct Threshold

[80] Reasonable suspicion guarantees the privacy protections that the law’s subjective “good faith” test does not because it provides the objectivity which the “good faith” test lacks. Unlike the “good faith” test, reasonable suspicion requires border officers to rely on objective facts supporting a possibility of border violations that courts can independently scrutinize. This prevents border officers from relying on mere hunches, intuition, and uncorroborated tips of unknown reliability. Likewise, it protects against Canfield’s concern that the state could use border search powers to conduct mass surveillance because officers cannot reasonably suspect everyone at a border crossing or arriving in Canada from a particular location: R. v. Ahmad, 2020 SCC 11, [2020] 1 S.C.R. 577, at paras. 46‑50; see Canfield, at para. 108. Similarly, sending travellers to secondary inspection does not establish reasonable suspicion because officers do not require grounds to do so: Hudson, at para. 35.

[81] The reasonable suspicion threshold also provides greater protection against the risk of discrimination. As the Supreme Court held, it is “designed to avoid indiscriminate and discriminatory searches”: Chehil, at para. 30. Its objectivity requirement achieves this goal by demanding facts showing a reasonable possibility of crime, not racial stereotypes: Chehil, at paras. 3, 25, 39, 42.

[82] At the same time, reasonable suspicion fairly accounts for the state’s important border control interests and travellers’ reduced expectation of privacy at the border. It does so by reducing the threshold from the highest threshold, reasonable and probable grounds, to an intermediate threshold. This matches the threshold for strip searches at the border, which are also among the most intrusive searches because they seriously infringe privacy and human dignity: Golden, at para. 99. It is also lower than the threshold for body cavity searches, which the Supreme Court selected as the apex Simmons category because, like the non-consensual seizure of bodily samples, they are the “ultimate affront to human dignity”: Stillman, at para. 39; see also at para. 42, citing Simmons, at p. 517; Fearon, at para. 59.

[83] I also agree with the trial judge that a reasonable suspicion threshold will not unduly interfere with the Agency’s mandate. The Agency does not need to meet that threshold to ask travellers questions and search their belongings, because those are Simmons category one searches that only require good faith purpose. As Mr. Pike submits, the information those routine questions and searches uncover permit the Agency to let innocent travellers pass while also establishing reasonable suspicion as to more concerning travellers that then justifies a digital device search. Indeed, as the trial judge reasoned, that is what the Agency did in Mr. Pike’s case. By investigating his background, asking him questions, and searching his luggage, Officer Grama learned that he had a criminal record for sexually abusing young boys, claimed to have been teaching boys in that same age group abroad despite his loss of his teacher’s license, was nervous and initially lied about his overseas activities, and was travelling with 14 personal digital devices and children’s items in his luggage—all of which taken together established reasonable suspicion. This provides comfort that a reasonable suspicion threshold will allow the Agency to target people who travel internationally to sexually exploit children without sweeping in numerous innocent travellers.

[84] While reasonable suspicion is the presumptive threshold for digital device searches, I do not wish to foreclose Parliament from adopting a lower threshold for certain discrete categories of information that may attract lower privacy interests. For instance, Canfield suggested that Parliament could set a lower standard to examine receipts, goods valuation documents, and travel-related documents contained on digital devices as part of routine screenings without limiting s. 8 rights, since these types of information engage lower privacy expectations: at para. 79. If Parliament chooses to create more targeted search powers to do so, it should take care to define the categories of information they target precisely and adopt safeguards to prevent border officers from using those powers to conduct more intrusive digital device searches without forming reasonable suspicion.

A Lower Threshold Is Inappropriate

[85] The Crown submits that digital device searches do not merit the reasonable suspicion threshold Simmons assigns to strip searches because they are less intrusive than those searches. This submission fails. It is based on Fearon’s comments that cell phone searches are “completely different from” and “not as invasive as [] strip search[es]”: at paras. 55, 63. But what Fearon actually held is that, while strip searches always gravely invade privacy, digital device searches vary in their intrusiveness: at paras. 54-55, 61. These subsequent comments, once placed in context, merely refer to that earlier, less categorical holding. They cannot be stretched to mean that any digital device search, even the most intrusive the law permits, is always much less intrusive than a strip search. That would contradict the Supreme Court’s pre- and post-Fearon holdings that digital device searches are among the most intrusive searches imaginable: Bykovets, at para. 54; Vu, at para. 40; Morelli, at paras. 2-3, 105. Further, it would undermine Fearon’s holding that digital device searches require meaningful protections because they often do cause “broad invasion[s] of privacy”: at para. 54. This holding applies here because, since a Simmons category provides the same protection to all searches it covers (Hudson, at paras. 26-30), the digital device search threshold must provide adequate protection for the most intrusive searches.

[86] Three additional factors confirm my conclusion that Fearon does not require a threshold lower than reasonable suspicion for digital device searches. First, while Fearon appears to have assumed that digital device searches only engaged informational privacy, the Supreme Court subsequently confirmed that viewing intimate images and videos, which travellers frequently store on their digital devices, also engages personal privacy and human dignity: Jarvis (2019), at paras. 62, 65, 67. Thus, both digital device searches and strip searches engage a similar dignitary interest in not being viewed in a state of undress without consent, which favours granting them a similar degree of protection: Doll, at p. 112. Second, even assuming that strip searches are moderately more intrusive than the most intrusive digital device search—a position I do not accept—Simmons category two can still accommodate both because it includes searches of varying intrusiveness: Monney (SCC), at para. 46. Third, as I explained, the state interest in preventing the entry of electronic contraband by searching digital devices is lower than that in preventing illegal drugs’ entry by strip searching travellers. This counterbalances any increased intrusiveness of strip searches relative to digital device searches.

[87] The Crown next submits that reasonable suspicion strikes the wrong balance because the trial judge found that the Agency did not meet this threshold as to Mr. Scott and it might permit other guilty travellers to evade detection. I disagree because, as I explain in the footnote at the end of this sentence, the facts the Agency knew about Mr. Scott establish reasonable suspicion, a question which is reviewed for correctness: Chehil, at para. 60.[^9] Moreover, the underlying logic of the Crown’s submission, that a threshold must be too high if it lets any guilty travellers evade detection, cannot be right. I agree with the trial judge that this submission’s one-sided approach fails to strike the fair balance that s. 8’s commitment to reasonableness requires and overlooks the other side of the ledger—the invasion of many innocent travellers’ digital privacy that the law’s low threshold enables. Indeed, it is inconsistent with Simmons’ decision to set a reasonable suspicion threshold for strip searches, since that also conceivably might let some travellers who are carrying drugs on their persons avoid detection.

[88] Further, the lower thresholds the Crown proposes simply do not cut it. Citing Canfield, the Crown first proposes Fearon-type attempts to tailor the search’s nature and extent to its purpose: at paras. 77-79. I do not wish to foreclose the possibility that Fearon’s approach may work for certain precisely defined types of information that are less private and are stored in specific locations on digital devices, such as boarding passes, receipts, and certain travel-related documents: Canfield, at para. 79. But this approach cannot work globally and is no substitute for reasonable suspicion as the presumptive threshold. As Canfield recognized, officers perform border searches in a “unique context” and for different purposes than non-border searches incident to arrest: at para. 79. That different context and purpose prevent those Fearon-type restrictions from taking reasonable suspicion’s place as the presumptive threshold. They are premised on the assumption that officers usually need only view recently sent or drafted communications to advance a limited set of purposes, namely ensuring officer and public safety and preserving or discovering evidence: Fearon, at paras. 75-78. But that assumption breaks down at the border because, as the trial judge reasoned, officers’ purposes are not always so limited but instead may extend to preventing the entry of any prohibited goods into Canada, no matter where they are located on a device. As I explained, Officer Sodhi’s searches in this case prove this because their sweeping breadth reflected the breadth of the file categories that are likely to contain child pornography.

[89] The same is true of the distinction between supposedly routine and less intrusive manual searches and non-routine, more intrusive forensic searches that many American cases draw. I agree with the trial judge that this distinction, which as the CCLA notes, has not been universally adopted in the still-developing American trial-level and intermediate appellate court caselaw (see, e.g., Smith, at p. 396), is artificial and has limited significance. After all, manual searches can still invade large amounts of highly private information and, if officers invest the time, can be almost as revealing as forensic searches. Further, officers could easily use manual searches as a backdoor to gain information that would meet the higher threshold to conduct forensic searches: Bingzi Hu, “Border Search in the Digital Era: Refashioning the Routine vs. Nonroutine Distinction for Electronic Device Searches” (2022) 49 Am. J. Crim. L. 177, at pp. 189-190.

(d) The Crown Has Not Justified Limiting Travellers’ Rights

[90] Laws that limit Charter rights infringe those rights and are unconstitutional if the state cannot justify the limits they impose under s. 1 of the Charter. This Charter provision permits the state to make laws that impose reasonable limits to Charter rights that can be demonstrably justified in a free and democratic society. To justify the law’s limit on travellers’ s. 8 rights, the Crown must show that it is prescribed by law, has a sufficiently important purpose, and is proportional. The limit is proportional if its means are rationally connected to its purpose, it is minimally impairing of Charter rights, and its salutary effects are proportional to its deleterious effects: R. v. Brown, 2022 SCC 18, 472 D.L.R. (4th) 459, at paras. 110, 126; Greater Vancouver Transportation Authority v. Canadian Federation of Students – British Columbia Component, 2009 SCC 31, [2009] 2 S.C.R. 295, at para. 48.

[91] I conclude that the law infringes s. 8 of the Charter and is unconstitutional because the Crown has not met its burden. The Crown fails to show that the law’s low good faith purpose threshold is necessary because Parliament set a higher threshold for other searches that raise similar privacy concerns, and the Agency itself admits that a less restrictive alternative, requiring border officers to rely on objective facts showing that travellers may be violating border laws, would not jeopardize its mandate.

Prescribed by Law

[92] To begin with, while s. 99(1)(a) itself is prescribed by law, I agree with the trial judge that the Agency’s policy guidance is not. Administrative policies designed to provide internal guidance to public officials do not meet this step because they are not intended to establish individual rights or authorize state action, but rather to help officials apply statutes and regulations: Greater Vancouver, at paras. 62-63. The Agency’s policy guidance fits that bill because, as the Vice-President of its Travellers Branch, Denis Vinette, testified, it aims to provide internal guidance to border officers rather than legally restrict their authority to search. The Crown thus cannot rely on that policy guidance to defend the law’s constitutionality.

Pressing and Substantial Purpose

[93] The law does advance a sufficiently important purpose, which I define as preventing the entry of prohibited electronic documents and unauthorized persons and ensuring the payment of duties on those documents or other goods they describe. Unlike the trial judge’s broader framing of the objective as “policing Canada’s borders and regulating who and what comes into the country,” this objective appropriately narrows Simmons’ discussion of the Customs Act’s general objectives to the rights-limiting measure’s specific target, electronic documents: Simmons, at pp. 527-528; Canada (Attorney General) v. JTI-Macdonald Corp., 2007 SCC 30, [2007] 2 S.C.R. 610, at para. 38.

Rational Connection

[94] I agree with the trial judge that the law is rationally connected to its objective. As Canfield held, by authorizing the Agency to search digital devices using a low threshold, it increases the likelihood of finding prohibited electronic documents: at paras. 86-90.

Minimal Impairment

[95] However, I agree with the trial judge that the next step, minimal impairment, is fatal to the law’s constitutionality and shows that the law infringes s. 8 of the Charter. Laws pass this step if they limit the right “as little as reasonably possible” to achieve the legislature’s purpose: Brown, at para. 135. They fail this step if alternative means that limit the Charter right less severely would allow the government to fully realize its objective: R. v. K.R.J., 2016 SCC 31, [2016] 1 S.C.R. 906, at paras. 70, 75-76. Courts can test the government’s assertions against its actions to determine this. For instance, the government could not justify an absolute liability offence’s limit on s. 7 rights because Parliament, by adopting a due diligence defence covering the same prohibited conduct, had all but admitted that the rights-limiting measure was not necessary: R. v. Hess, 1990 CanLII 89 (SCC), [1990] 2 S.C.R. 906, at pp. 925-926; see also Goodwin, at paras. 84-85. In my view, courts may draw the same conclusion when the agency Parliament tasks with administering a legislative scheme concludes that less restrictive alternatives are adequate to advance its mandate.

[96] The law’s low threshold for searches of digital devices is not minimally impairing because it is lower than higher thresholds that Parliament has deemed adequate in the border context for similarly intrusive searches. While Canfield listed many examples (at paras. 98-99), s. 99(1)(b)’s adoption of a reasonable suspicion threshold for imported physical mail sufficiently proves the point. Like the electronic mail and messages that digital devices contain, physical mail carries high privacy interests because it often contains sensitive information concerning the sender or recipient’s identity, personal life, or political or religious beliefs: R. v. Crane (1985), 1985 CanLII 5997 (NL SC), 53 Nfld. & P.E.I.R. 82 (N.L. Dist. Ct.), at para. 29; see Branch, at para. 62. Further, physical mail was a major means of importing child pornography when Parliament enacted s. 99(1)(b) in 1985, just as digital devices are today.[^10] By choosing a reasonable suspicion threshold for physical mail searches, Parliament recognized that providing meaningful privacy protections would not frustrate its objective of preventing the entry of child pornography and other prohibited items. The same is true of digital devices. They merit at least the same protection because people who travel with them are typically carrying every piece of electronic mail they recently sent or received.

[97] The Agency executive’s testimony that less restrictive alternatives would adequately advance the Agency’s mandate confirms that the law fails this step. Having a search threshold that requires officers to point to objective facts that show a possibility of border law violations impairs privacy less than the law’s present, purely subjective good faith purpose threshold. Both pre- and post-Canfield, the Agency’s policy guidance instructs officers to point to those very objective facts suggesting the possibility of border law violations that the law itself does not require before searching digital devices: Canfield, at paras. 47, 96. Vice-President Vinette testified that he expected Agency officers to follow this policy guidance, it was functioning well, and it did not jeopardize the Agency’s ability to fulfill its mandate. While the trial judge criticized this policy guidance as vague, discretionary, and insufficiently protective of privacy, it still imposes a higher threshold than the law’s purely subjective test by requiring officers to point to at least some objective facts. As Mr. Scott submits and Canfield held, these self-imposed policy restrictions show that enshrining them in law would not frustrate the Agency’s mandate of regulating the cross-border movement of people and goods: at para. 97.

[98] Vice-President Vinette’s stated preference to enshrine limits in policy rather than law does not change this conclusion. Of course, the Agency might prefer internal accountability mechanisms to a legal threshold, as he testified. But as the trial judge reasoned, it is precisely the independent judicial scrutiny that a legal threshold provides that travellers need to protect against arbitrary action by border officers: see Chehil, at paras. 45-47. Whatever its preferences, the Agency can rise to the challenge of complying with the Charter and exercising its extensive powers in a manner that will stand up in court: Le, at para. 165.

Proportionality of Effects

[99] While my conclusion on minimal impairment establishes the law’s unconstitutionality, I agree with the trial judge that it also fails the proportionality of effects step because its harmful effects are disproportionate to its benefits. At this step, courts normatively assess both the marginal benefits the rights-limiting measure provides over less restrictive alternatives and the degree the law impairs the right’s purposes: K.R.J., at paras. 77, 89-90; Alberta v. Hutterian Brethren of Wilson Colony, 2009 SCC 37, [2009] 2 S.C.R. 567, at para. 88. Here, the law’s marginal benefits are meagre at best because the Agency’s current policy guidance, which adopts a modest threshold, does not prevent it from achieving its border control and regulatory mandate. In contrast, the law’s harms are severe. Because digital device searches are among the most intrusive imaginable, they cut at the heart of s. 8’s purpose of protecting privacy and, in turn, its related goals of fostering dignity, autonomy, personal growth, and democratic flourishing: Bykovets, at para. 29. Such searches invade not merely informational privacy but also personal privacy and human dignity. Further, the low threshold increases the risks of racial profiling and the searching of innocent travellers. It is not reasonable or justifiable in a free and democratic society to invade the most vital privacy interests of so many innocent people for so little demonstrable benefit.

(iv) Remedy: The Law’s Authorization of Digital Device Searches Is of No Force or Effect

[100] Determining that a law is unconstitutional does not end the court’s work. Instead, it begins a multi-stage process to determine how to remedy the unconstitutionality and provide effective remedies for the claimant and other people whose rights the unconstitutional law violates. The starting points of this analysis are s. 52(1) of the Constitution Act, 1982, which provides that laws are of no force or effect to the extent they are inconsistent with the Constitution’s provisions, and s. 24(1) of the Charter, which authorizes anyone whose Charter rights or freedoms have been violated to seek an appropriate and just remedy.

[101] The Constitution’s remedy provisions require us to make up to five decisions. First, we must determine the extent to which the law is unconstitutional and whether to declare the entire law inoperative under s. 52(1) or grant a narrower remedy that is tailored to the breadth of the violation: Ontario (Attorney General) v. G., 2020 SCC 38, [2020] 3 S.C.R. 629, at para. 108, 112. Second, we must decide whether to suspend the declaration of unconstitutionality, meaning to delay it from taking effect immediately, and, if so, for how long: G., at paras. 117, 134-135. Third, we must resolve whether the declaration should be prospective or retrospective, meaning whether the law is only inoperative from the time of the declaration onwards or is instead inoperative from the time it was enacted: R. v. Albashir, 2021 SCC 48, 464 D.L.R. (4th) 199, at paras. 44-53, per Karakatsanis J., and at para. 84, per Rowe J. (dissenting but not on this point). Fourth, if we issue a suspended declaration, we must choose whether, as a s. 24(1) Charter remedy, to exempt the claimant from it, meaning that the law will be treated as of no force and effect as applied to the claimant: G., at paras. 140-152. Finally, if we issue a suspended and prospective declaration, we should provide guidance on the circumstances in which other people, who were subject to the unconstitutional law before it was declared unconstitutional or during the suspension period, can seek a s. 24(1) remedy: Albashir, at paras. 62-71.

[102] It is unfortunate that the parties did not make submissions concerning the last four of these decision points. They implicate important public interests concerning effective Charter remedies, constitutional compliance, the public’s entitlement to the benefit of laws that the legislature passed to advance the public good, and respect for the legislature’s democratic legitimacy and law-making role: G., at paras. 94-99. Courts’ decisions concerning them can have profound impacts, both on those whose rights the law violates and those whose interests the legislature sought to protect. Litigants, especially government litigants who are responsible for defending the public interest, should not gloss over these questions and must come to court with a clear position that, where possible, is evidence-based. We nonetheless determined that the parties’ arguments concerning the merits of the constitutional challenge enabled us to resolve these remedy questions.

[103] First, I would read down the law by declaring that its authorization to search the electronic documents that digital devices contain is of no force and effect. Reading down involves declaring the law inoperative only “to a precisely defined extent”: G., at para. 113. This remedy is appropriate if the law’s unconstitutional portion can be defined in a limited manner and Parliament would have enacted the law without that portion: at paras. 113, 116. Both conditions are met. Only the law’s authorization to search the electronic documents that digital devices contain is unconstitutional, not its authorization to search other goods, which Mr. Pike and Mr. Scott have not challenged. Further, because the law serves important border control purposes for physical goods, Parliament would have enacted the law to advance those purposes even without the offending electronic documents portion. This is similar to Canada (Attorney General) v. Federation of Law Societies of Canada, where the Supreme Court remedied the unconstitutionality of broad examination provisions only infringed s. 8 of the Charter to the extent they authorized searches of legal offices by excluding those searches from the provisions’ scope: 2015 SCC 7, [2015] 1 S.C.R. 401, at paras. 62-63.

[104] I would not read-in a reasonable suspicion standard to the law because I cannot be sure that Parliament would have done so. Parliament could also adopt more privacy protective thresholds, adopt different thresholds for different types of information, add additional oversight or reliability guarantees, incorporate intrusiveness limits like s. 99.3(1)’s customs controlled area search power, provide for prior judicial or administrative authorization requirements as in Simmons or, perhaps, attempt to defend some other threshold higher than subjective good faith purpose but lower than reasonable suspicion as a justified limit under s. 1: Canfield, at para. 114. Parliament gets to decide which of those options to choose to correct the law’s constitutional defects: Federation of Law Societies of Canada, at para. 65.

[105] Second, I would suspend the declaration of unconstitutionality and delay it from coming into effect by six months. The government has the onus to prove that a suspended declaration is necessary to protect its ability to legislate and set policy. To meet this onus, the government must show that the law advances a compelling public interest that outweighs the importance of providing effective remedies for Charter violations and the public’s interest in constitutionally compliant legislation: G., at paras. 129-133, 139.

[106] The same factors the Crown relied on to defend the law’s constitutionality justify a modest extension. Like the border search law Simmons considered, subsection 99(1)(a)’s authorization of digital device searches advances “compelling” and “important state interests”: at p. 528. Immediately declaring this law unconstitutional would undermine those interests and public safety by preventing the Agency from searching travellers’ digital devices, even where the facts established reasonable suspicion: Canfield, at para. 113; see also R. v. Gorman, 2022 NLSC 67, at paras. 16-18. This would jeopardize the state’s ability to protect children from sexual abuse and exploitation committed by perpetrators like Mr. Scott and, the Crown alleges, Mr. Pike. I agree with the Court of Appeal for British Columbia that courts should be loath to deprive the state of the necessary tools to protect children from these abhorrent crimes: T.L. v. British Columbia (Attorney General), 2023 BCCA 167, 481 D.L.R. (4th) 65, at paras. 280-284; see also R. v. Ndhlovu, 2022 SCC 38, 474 D.L.R. (4th) 389, at para. 139. While granting a suspension temporarily prolongs the rights violation and undercuts the public’s interest in constitutional compliance, the Agency’s policies somewhat temper these concerns by directing border officers to only search digital devices if they have an objective basis to do: Gorman, at para. 21. For all these reasons, Parliament should have the chance to make an informed choice from the multiple policy options open to it: Reference re Code of Civil Procedure (Que.), art. 35, 2021 SCC 27, [2021] 2 S.C.R. 291, at para. 154.

[107] The Court of Appeal for Alberta’s decision not to extend the suspended declaration it granted in Canfield does not change this conclusion. After granting a six-month extension to the initial one-year suspended declaration it issued in October 2020 (2021 ABCA 352, 407 C.C.C. (3d) 27), that court ruled 2-1 to deny a second extension because, in the majority’s view, the federal government had not made the case that it needed more than 18 months to legislate: 2022 ABCA 145, 45 Alta. L.R. (7th) 38. But because Canfield does not bind this court and its declaration of unconstitutionality has no effect in Ontario (R. v. Sullivan, 2022 SCC 19, 472 D.L.R. (4th) 521, at para. 62), the Crown was entitled to defend the law’s constitutionality in the Ontario courts. Parliament should thus be given a chance to legislate with the benefit of this court’s decision, which differs meaningfully from Canfield.

[108] However, I would limit the suspension to six months. This moderate suspension is appropriate because, following Canfield, Parliament began considering draft legislation to address digital device searches at the border, Bill S-7, “An Act to Amend the Customs Act and Preclearance Act, 2016,” which has cleared the Senate and completed first reading in the House of Commons. This “head start” eliminates the need for a longer suspension, which the Crown has not sought to justify: Bjorkquist v. Canada (Attorney General), 2023 ONSC 7152, 169 O.R. (3d) 1, at para. 283.

[109] Third, I would also make the suspended declaration prospective, meaning that the law will only be inoperative on a forward-looking basis once the declaration takes effect, and not from the moment the law was enacted in 1985. A prospective declaration is necessary to advance the same border regulation, public safety, and child protection considerations that require a suspended declaration: Albashir, at para. 46; T.L., at para. 287.

[110] Fourth, I would exempt Mr. Pike and Mr. Scott from the suspended declaration so that the law is of no force or effect in the Crown’s case against them. The Supreme Court has held that litigants who successfully challenge a law’s constitutionality are generally entitled to an exemption because they have advanced the public interest in constitutional compliance: G., at paras. 147-148, 152; Ndhlovu, at para. 141. This general rule applies here because a suspension would not unduly jeopardize the interests that the law advances: G., at para. 151. This is so because, as I explain later in these reasons, it does not automatically lead to the exclusion of the evidence of the child pornography files that the Agency found on their digital devices. Rather, the test for exclusion of evidence under s. 24(2) of the Charter accounts for the state’s interest in holding Mr. Pike and Mr. Scott accountable for the child pornography crimes with which they were charged, as well as for border officers’ good faith reliance on the law’s constitutionality.

[111] Finally, courts have discretion to grant personal remedies under s. 24(1) of the Charter to people who were searched under the law before this court’s decision or during the period of the suspension. They can grant those remedies if: (1) a search triggers the constitutional defects this decision identified, meaning that the officer(s) conducting the search lacked reasonable suspicion; and (2) granting a remedy would not undermine the purpose of the suspension: Albashir, at paras. 67-69, 71. This means that the Agency should only search travellers’ digital devices during the suspension period if its officers have reasonable suspicion to justify the search. This enables the Agency to use the law to temporarily target those travellers whom it could legitimately search under a constitutionally compliant law without continuing to sweep in innocent travellers whom it has no grounds to suspect: at para. 69.

(2) The Agency Violated Mr. Pike’s and Mr. Scott’s Right to Counsel

[112] Sections 10(a) and (b) of the Charter protect the right of detained people to understand the jeopardy they face and to speak to counsel by requiring the state to inform them of the reasons for their detention promptly and, without delay, of their right to counsel. Whether the Agency breached that right turns on when it detained Mr. Pike and Mr. Scott. The trial judge found that the Agency detained them when it demanded their digital device passwords, breached their s. 10(b) rights by failing to inform them then of their right to counsel without delay, and also breached Mr. Scott’s s. 10(a) rights by failing to inform him of the reasons for his detention. I agree. In this section of the reasons, I explain how the law of detention functions at the border and when and why the Agency detained Mr. Pike and Mr. Scott and breached their rights to counsel and, in Mr. Scott’s case, his right to be informed of the reasons for his detention.[^11]

(a) Detention Law at The Border

[113] The Charter protects people’s liberty from unjustified state interference by granting each person who is detained certain specified rights. Detentions trigger both the s. 9 Charter right not to be arbitrarily detained and s. 10 Charter rights, including the s. 10(a) right to be informed promptly of the reasons for the detention and the s. 10(b) right to counsel. These rights remedy the power imbalance between the state and detained people by requiring the state to inform detained people of the jeopardy they face and their right to speak to a lawyer, and to permit them to exercise that right. They ensure that detained people can seek legal advice concerning how to regain their liberty and make an informed choice whether or not to speak to authorities: Le, at para. 29; R. v. Grant, 2009 SCC 32, [2009] 2 S.C.R. 353, at paras. 21-22; R. v. Latimer, 1997 CanLII 405 (SCC), [1997] 1 S.C.R. 217, at para. 28.

[114] Outside the context of border detentions, the state detains people when it removes their choice to walk away from state authorities by physical or psychological compulsion. Psychological compulsion means a demand or direction by a state agent to an individual, which a reasonable person in the individual’s position would feel obligated to comply with and conclude that they are not free to leave: Grant, at para. 21; Le, at paras. 25‑26.

[115] Canada’s interest in controlling and regulating its border, coupled with the reduced stigma of border searches, increase the threshold for detention at the border relative to non-border contexts. Under this higher threshold, routine border questions and search demands do not trigger a detention. This is so even though these questions and demands would likely trigger a psychological detention outside the border context, as travellers are legally required to comply and are not free to leave and enter Canada until they do so: Customs Act, ss. 11(1), 13, 99.1. More is required at the border: Simmons, at pp. 517, 521; Dehghani v. Canada (Minister of Employment and Immigration), 1993 CanLII 128 (SCC), [1993] 1 S.C.R. 1053, at pp. 1068, 1071-1074; R. v. Barac, 2023 ONCA 216, 425 C.C.C. (3d) 40, at paras. 18-19.

[116] This court has adopted a two-branch test to define the higher threshold for detention at the border. The first branch, which I refer to as the intrusive search branch, provides that non-routine, more intrusive questioning and searches trigger a detention. This branch is met by: (1) Simmons categories two and three searches, meaning strip searches, body cavity searches, and now, digital device searches; and (2) questions that contain improper inducements, exert unfair pressure, or involve coercive or adversarial interrogation. The second branch, which I refer to as the reasonable expectation branch, considers three detention factors, namely whether officers: (1) have subjective particularized suspicion that a traveller has violated border laws, (2) subjectively decide to go beyond routine questioning and engage in a more intrusive inquiry, and (3) take actions that cause a reasonable person in the traveller’s position to expect that they would be subjected to a non-routine, more intrusive inquiry: Barac, at paras. 20-21; R. v. Ceballo, 2021 ONCA 791, 408 C.C.C. (3d) 70, at paras. 20-30.[^12]

[117] I wish to clarify two points concerning this two-branch test.

[118] First, branch two, the reasonable expectation branch, is an objective test. Its final, objective factor is both necessary and sufficient to establish a detention. Thus, a traveller is detained if an officer’s conduct causes a reasonable person in the traveller’s position to expect that they are the subject of a non-routine and more intrusive inquiry that would trigger branch one, the intrusive search branch. This is so even if the first two subjective factors, the officer’s particularized suspicion and decision to conduct a non-routine, more intrusive search, are absent. This objective test furthers the rule of law by subjecting all claims to the same standard, regardless of officers’ or travellers’ subjective perceptions: Le, at para. 115. The first two subjective factors, while not necessary, make it more likely that the final factor’s objective test is met by supporting an inference that officers followed through on their subjective suspicion and decision to search by their actions: Grant, at paras. 41, 49.

[119] I reach this conclusion because the Supreme Court and this court have long held that the test for psychological detention at the border is objective: Dehghani, at pp. 1066-1068, citing R. v. Kwok (1986), 1986 CanLII 4726 (ON CA), 31 C.C.C. (3d) 196 (Ont. C.A.), at p. 207. This means that an officer’s subjective suspicions and intentions are neither necessary nor sufficient to establish a detention: R. v. Lafrance, 2022 SCC 32, 472 D.L.R. (4th) 1, at para. 27; Le, at para. 37; Grant, at para. 32. For example, an officer’s “[f]ocussed suspicion [of the accused], in and of itself, does not turn the encounter into a detention” absent action that would cause a reasonable person in the accused’s position to believe that the accused is detained: Grant, at para. 41. Neither does an officer’s intent to detain and search that is unaccompanied by action: R. v. MacMillan, 2013 ONCA 109, 114 O.R. (3d) 506, at para. 37. Conversely, officers can unintentionally detain a person they do not suspect if their actions cause a reasonable person in that person’s position to conclude that the person is detained: R. v. Johns (1998), 1998 CanLII 2667 (ON CA), 123 C.C.C. (3d) 190 (Ont. C.A.), at para. 27; R. v. Tutu, 2021 ONCA 805, 407 C.C.C. (3d) 137, at paras. 15, 19-21; Lafrance, at paras. 35-40.

[120] My conclusion is consistent with this court’s 2006 decision in Jones, because it implicitly affirmed the objective test that I expressly adopt. Jones did so by following R. v. Jacoy, 1988 CanLII 13 (SCC), [1988] 2 S.C.R. 548, which held that “the trigger for detention was not … the [officer’s] subjective intention …, but rather the steps … taken to commence the intrusive investigation,” which caused a reasonable person in the traveller’s position to conclude that he was detained: Ceballo, at para. 29; see Jones (2006), at para. 42. Further, the implied requirement that officers must act for a good faith purpose that I previously described addresses Jones’ concern about using border search powers to investigate crime by preventing officers from using those powers as a ruse or pretext to do so: Mayor, at paras. 7-9; Appleton, at para. 12. For those reasons, and because Jones used conditional language to describe the reasonable expectation branch, I do not take its reference to the first two subjective factors to mean that they are necessary elements: Jones (2006), at para. 42.

[121] Second, the trial judge erred by suggesting that demands with significant legal consequences are a third detention branch. Rather, such demands are a means of satisfying the final, objective factor of branch two, the reasonable expectation branch: Ceballo, at para. 29. This makes sense because both the current Customs Act and its predecessor that Simmons considered make it an offence to resist or refuse to comply with any lawful border search, whether routine or non-routine: Customs Act, ss. 153(c), 153.1, 160-160.1; Simmons, at p. 521. The Supreme Court held that, for routine searches, these legal consequences do not trigger a detention because, without more, they do not cause a reasonable person in the traveller’s position to expect to undergo a non-routine, intrusive search. In contrast, if border officers demand that travellers submit to a non-routine, intrusive search, then these legal consequences confirm that there is a reasonable expectation branch detention. By preventing travellers from leaving or resisting the demands without committing an offence, these legal consequences cause travellers to reasonably expect that they must comply with the demands and will undergo a non-routine, intrusive search: Simmons, at pp. 517, 521; Jacoy, at pp. 557-558; Dehghani, at pp. 1071-1074.

(b) The Agency Delayed Providing the Right to Counsel

[122] I agree with the trial judge that the Agency detained Mr. Pike and Mr. Scott when its officers demanded their digital device passwords. Because digital device searches, like strip searches, are Simmons category two non-routine searches, these demands triggered reasonable expectation branch detentions like the strip search demand in Simmons. The password demands would cause a reasonable person in Mr. Pike’s and Mr. Scott’s position to expect that the Agency would subject them to a non-routine, more intrusive search. I reject the Crown’s challenge to the detention finding because it is based on the incorrect premise that digital device searches are routine Simmons category one searches. Further, while the trial judge’s reasons also discussed the presence or absence of officers’ subjective suspicion and intent to search digital devices, these factors are neither necessary nor sufficient to establish a detention for the reasons I have explained.

[123] Because the demands triggered detentions, I also agree with the trial judge that the Agency breached Mr. Pike’s and Mr. Scott’s s. 10(b) right to counsel. Instead of informing them of that right upon their detention without delay, the Agency waited 1 hour and 24 minutes, and 55 minutes, respectively. For the same reason, the Agency breached Mr. Scott’s s. 10(a) right to be informed promptly of the reasons for his detention.

(3) The Evidence Against Mr. Pike and Mr. Scott Is Admissible

[124] Section 24(2) of the Charter provides that evidence obtained in a manner that breaches Charter rights shall be excluded if its admission would bring the administration of justice into disrepute. Courts evaluate whether to exclude evidence by balancing the three Grant factors: (1) the seriousness of the Charter‑infringing state conduct, (2) the impact of the breach on the accused’s Charter‑protected interests, and (3) society’s interest in trying the case on its merits: Grant, at para. 71. While appellate review is deferential, this deference is displaced and the appellate court can conduct a fresh analysis if the trial judge overlooks or disregards relevant factors, errs in law or principle, or makes an unreasonable determination: R. v. Szilagyi, 2018 ONCA 695, 142 O.R. (3d) 700, at para. 41. The Crown can appeal these legal errors: R. v. Harris, 2007 ONCA 574, 87 O.R. (3d) 214, at para. 51. Even if the trial judge committed a legal error and should have admitted the evidence, the Crown still has a heavy burden to overturn an acquittal. It must show to a reasonable degree of certainty that the verdict would not necessarily have been the same but for the legal error: R. v. Hodgson, 2024 SCC 25, at para. 36.

[125] The trial judge excluded the evidence the authorities found on Mr. Pike’s digital devices under s. 24(2) but admitted the evidence from Mr. Scott’s devices. I agree with the Crown that the trial judge erred in excluding the evidence against Mr. Pike but would defer to his decision to admit the evidence against Mr. Scott.

(a) The Trial Judge Should Have Admitted the Pike Evidence

[126] I would admit the evidence of the child pornography files found on Mr. Pike’s devices against him because, in my respectful view, the trial judge overstated the seriousness of the Agency’s conduct and overlooked factors that reduce the impact of the Charter breaches on Mr. Pike and bolster society’s interest in a trial on the merits. These legal errors caused him to exclude the child pornography files that were critical to the Crown’s case and, in turn, to acquit Mr. Pike. The Crown has thus met its heavy burden to show that the verdict would not necessarily have been the same but for these legal errors, as in R. v. Mackey, 2020 ONCA 466, 392 C.C.C. (3d) 230, at paras. 77-80, leave to appeal refused, [2020] S.C.C.A. No. 293. I would accordingly overturn the acquittal and order a new trial.

[127] First, the trial judge made errors of law that caused him to overstate the seriousness of the Charter-infringing state conduct. Charter breaches are in good faith and, thus, less serious, if officers were acting consistently with what they reasonably believed to be the law: Le, at para. 147. The trial judge accepted that Officer Grama breached s. 8 in good faith because the caselaw at the time of the search established that the search was valid. However, he characterized the s. 10(b) breach as flagrant because, in his view, Officer Grama should have known from the pre-2020 Jones and Jacoy cases that his subjective suspicion of Mr. Pike, intent to search all his devices, and password demand all effected a detention. I do not agree that those cases establish this conclusion. As I explained, this court’s 2006 Jones decision does not depart from the objective detention test, which establishes that an officer’s subjective suspicion and intent are insufficient to trigger a detention. Likewise, while Jacoy held that demands to undergo a non-routine search will trigger a detention, it did not establish that digital device searches are non-routine or that demands for those devices’ passwords trigger detentions. The first case to do so, Canfield, both postdated the search of Mr. Pike’s devices and departed from the prior caselaw: at paras. 69, 163, 165.

[128] On a fresh assessment, the first Grant factor is neutral. Because the caselaw at the time of the search supported Officer Grama’s actions and he could not have foreseen subsequent court rulings, both the ss. 8 and 10(b) breaches were made in good faith, which favours admission: Grant, at para. 75. At the same time, I agree with the trial judge that Officer Grama’s breach of the Agency’s note-taking policy and misleading trial testimony about his subjective suspicion of Mr. Pike were serious. His misleading testimony significantly favours exclusion because it “directly undermine[s]” the judicial system’s integrity and truth-seeking function: R. v. Harrison, 2009 SCC 34, [2009] 2 S.C.R. 494, at para. 26 (quotation omitted). But its impact on the seriousness of the state conduct is reduced because it concerned a non-essential subjective detention factor rather than Officer Grama’s actual actions, which matter most to the objective detention test and were consistent with the caselaw’s then-dominant view: R. v. Dosanjh, 2022 ONCA 689, 163 O.R. (3d) 401, at paras. 175-178. Overall, the false testimony and note-taking failure counterbalance Officer Grama’s good faith and make the first Grant factor neutral.

[129] Second, the trial judge overlooked factors that reduce the impact of the ss. 8 and 10(b) breaches on Mr. Pike. The trial judge characterized the breach of Mr. Pike’s s. 8 Charter right as extremely serious. But it is an error in principle to fail to consider a reasonable expectation of privacy’s diminished nature: Cole, at para. 92. That error taints the trial judge’s conclusion because he overlooked that the border context diminished Mr. Pike’s privacy interest in his digital devices: R. v. Al Askari, 2021 ABCA 204, 406 C.C.C. (3d) 503, at para. 127. I would instead characterize it as “significant,” as in Canfield: at para. 174. Similarly, the trial judge overlooked that the s. 10(b) breach was a relatively moderate delay of 1 hour and 24-minutes: R. v. Griffith, 2021 ONCA 302, 408 C.C.C. (3d) 244, at paras. 72-75. This error undermines his conclusion that the breach was “at the most serious end of the spectrum” because those cases involve delays of several hours or more: Griffith, at para. 73. I would instead characterize the breach’s impact as significant because, while its duration was moderate, the trial judge found that it was causally linked to the evidence which the Agency obtained.

[130] Third, the trial judge also overlooked several key factors in this case which increase society’s interest in an adjudication on the merits. His comment that this factor “always” favours inclusion shows that he did not account for the child pornography files’ enhanced reliability and importance in this case. These factors favour admission because the files are reliable real evidence whose exclusion gutted the Crown’s case: Grant, at paras. 81-83; R. v. Spencer, 2014 SCC 43, [2014] 2 S.C.R. 212, at para. 80.

[131] Further, the trial judge mentioned the possibility that the seriousness of the child pornography offences with which Mr. Pike was charged might favour admission but ultimately did not resolve this question. To be clear, it did.

[132] Mr. Pike, a former teacher who lost his teaching licence after he was convicted for sexually abusing eight young boys and making child pornography recording that abuse, told Officer Grama that he was teaching seven-year-old boys for months in a country that people frequently visit to sexually exploit children. Many of the children depicted in the nearly 1,600 child sexual exploitation images stored on the digital devices that he brought back to Canada appeared to be the same age as the children he claimed to be teaching. Some of those images showed sleeping children, children in a bedroom with Mr. Pike, and a child sitting on his lap. Because the offences with which he was charged “‘implicate[] the safety of children,’” their seriousness strongly favours admission: Canfield, at para. 182, quoting Spencer, at para. 80. The seriousness of the offence does not cut both ways here because, unlike in R. v. Reeves, the Agency did not breach clearly established legal standards: 2018 SCC 56, [2018] 3 S.C.R. 531, at paras. 62-64, 67-68. Rather, as in Fearon, the caselaw at the time supported Officer Grama’s view that his actions were lawful: at paras. 92-95.

[133] Finally, the trial judge did not consider the offences’ international dimensions, which flow from and confirm the Crown’s argument that child pornography’s harmfulness to children is a serious societal concern that favours admission. By teaching children abroad and returning with devices that contained numerous child sexual exploitation images, Mr. Pike engaged Canada’s strong interests in “protect[ing] children [abroad] who are abused by Canadians” (R. v. A.V., 2022 ONCJ 328, at para. 24 (quotation omitted), aff’d, 2024 ONCA 339, 437 C.C.C. (3d) 316), and preventing travellers from misusing “the advantages of their Canadian nationality and residence” to do so with impunity: R. v. Klassen, 2008 BCSC 1762, 240 C.C.C. (3d) 328, at paras. 94-95. These interests are especially strong because they engage Canada’s international legal duties under Articles 3-4 and 10 of the Rights of the Child Protocol[^13] to cooperate with the countries that Canadian travellers visit to suppress child pornography crimes they commit against children of those countries. The countries Mr. Pike visited, nearly all of which are also parties to this treaty, count on Canada to uphold its treaty obligations by prosecuting alleged Canadian perpetrators because they have limited ability to apprehend and prosecute transitory visitors like Mr. Pike: Klassen, at paras. 21, 56, 95-97.

[134] On balance, I would admit the evidence and order a new trial because doing so does not bring the administration of justice into disrepute. Society’s interest in a trial on the merits is decisive because the first and second Grant factors, taken together, make only a moderate case for exclusion: R. v. Beaver, 2022 SCC 54, 475 D.L.R. (4th) 575, at paras. 134-135. As in Spencer, the evidence is highly reliable, its exclusion would gut the Crown’s case, and the offences implicate the safety of children. While those factors on their own might not be enough to overcome Officer Grama’s false testimony, the offences’ international dimensions confirm the case for admission. Society’s interest in ensuring accountability is even more powerful where, as here, Canada has both a strong interest and duty to hold Canadians accountable for allegedly sexually exploiting children during travel to countries that have a limited opportunity to prosecute them. The reasonable person concerned with the justice system and rule of law’s long-term integrity would be concerned about sending a message that alleged perpetrators can find safe haven in Canada and avoid answering the charges: Grant, at paras. 67, 79; Klassen, at paras. 94-97.

[135] I am also convinced that this decision will mitigate any negative impact on the justice system’s reputation that admitting the evidence might cause by establishing clear border search and detention standards for the Agency to follow. By doing so, the decision “render[s] similar conduct less justifiable going forward” and sends a message that the Charter rights to counsel and to be free from unreasonable searches matter at the border: Grant, at para. 133. This court expects that the Agency will rigorously comply with these standards because failing to honour the rule of law jeopardizes its legitimacy and ability to perform its important border security responsibilities: Le, at paras. 162, 165. Failing to do so may well result in the exclusion of evidence in future cases.

(b) The Trial Judge Reasonably Admitted the Scott Evidence

[136] In my view, the trial judge reasonably determined that admitting the evidence against Mr. Scott would not bring the administration of justice into disrepute. I would accordingly dismiss Mr. Scott’s appeal from conviction.

[137] First, the trial judge reasonably determined that the Agency searched Mr. Scott’s digital devices and delayed informing him of the reasons for his detention and of his right to counsel in good faith. He found that the caselaw at the time supported Officer Sodhi’s actions and that he could not have foreseen subsequent court rulings.

[138] I disagree with Mr. Scott’s argument that the good faith finding was unreasonable because, in his view, the Agency did not follow its own digital device search policy’s “progressive” search requirement. Because law enforcement policies are guidelines not law (Beaudry, at paras. 45-46), breaching them does not, without more, breach the Charter. While deliberate or negligent failure to follow policies designed to protect Charter rights can increase a Charter breach’s seriousness (R. v. Singh, 2016 ONCJ 386, 361 C.R.R. (2d) 138, at para. 19), that is not what happened here. Progressive examination means intensifying the level of examination as more indicators suggest a border law violation. Officer Sodhi and an officer who assisted him testified that they tried to follow this policy guidance. The trial judge did not find that they breached it. Instead, he found that the policy was vague and gave officers wide discretion. Given this, I am not surprised that the officers and Mr. Scott reasonably disagree about whether they complied with the policy, and resolving this disagreement has no bearing on the s. 8 breach’s seriousness.

[139] Mr. Scott’s remaining challenges to the trial judge’s good faith finding also fail. To begin with, I defer to the trial judge’s finding that Officer Sodhi’s failure to take notes was aggravating and decline to recharacterize that failure as even graver: R. v. Hamouth, 2023 ONCA 518, 167 O.R. (3d) 682, at paras. 36-37, leave to appeal refused, [2023] S.C.C.A. No. 387. Further, Officer Sodhi did not misunderstand the purpose of the right to counsel. Instead, he accurately testified that its purpose is to enable Mr. Scott to speak to a lawyer who could tell him that he does not need to speak to the authorities. Finally, unlike in R. v. Thompson, 2020 ONCA 264, 62 C.R. (7th) 286, here the Agency did not systematically violate established constitutional duties. Rather, at the time Officer Sodhi searched Mr. Scott’s devices, the Agency did not tell officers to provide the right to counsel before conducting digital device searches because the caselaw then held that those searches did not trigger that right. After Canfield overturned that view, the Agency course-corrected by adopting a new right to counsel policy that complies with Canfield’s guidance.

[140] Second, the trial judge reasonably found that the digital device search’s impact on Mr. Scott’s privacy, while serious, was less serious than in Pike. Mr. Scott challenges this finding, arguing that Officer Sodhi intended to search all the digital devices and searched many of them. I reject this challenge. Officer Sodhi and his colleagues did not search any of Mr. Scott’s eight hard drives. Instead, they only searched the three cell phones, one of the two tablets, and some USB keys—significantly fewer devices than in Mr. Pike’s case, where Officer Grama searched all 11 working devices. Officer Sodhi’s unrealized intent to search more devices is not relevant to the second Grant factor, which focuses on the breach’s “actual[]” impact on the right, not its hypothetical impact in an alternate scenario: at para. 76.

[141] Third, after considering the strong interest of society in a trial on the merit, the trial judge balanced the three Grant factors and concluded that Officer Sodhi’s good faith tipped the scales in favour of admission. I defer to this determination and reject Mr. Scott’s submission that the trial judge did not consider the breaches cumulatively. The trial judge expressly considered the seriousness of the state conduct underlying all the breaches together and assessed those breaches’ combined impact on Mr. Scott’s Charter-protected interests by reasoning that the right to counsel breach had a greater impact than the unreasonable search breach. Mr. Scott has not shown any reversible error.

(4) Sentencing Mr. Scott For Possessing Child Pornography

[142] The trial judge imposed a 23-month conditional sentence on Mr. Scott for importing child pornography and held that a prison term was unnecessary. The Crown seeks leave to appeal that sentence and argues that the trial judge erred by understating the seriousness of Mr. Scott’s conduct, and that he should have imposed a three-year prison sentence. I agree. Below I explain the governing sentencing principles, how the trial judge erred, the sentence he should have imposed, and why I would grant the Crown leave to appeal but dismiss the appeal to ensure that Mr. Scott serves the remainder of his conditional sentence.

(a) Sentencing Principles for Child Pornography Possession

[143] Ever since Parliament created the child pornography offences in s. 163.1 of the Criminal Code in 1993, courts have been on a “learning curve” to understand their wrongs and harms: R. v. D.G.F., 2010 ONCA 27, 98 O.R. (3d) 241, at para. 21. To further judicial progress along this learning curve, I apply the child-centered approach from R. v. Friesen, 2020 SCC 9, [2020] 1 S.C.R. 424, and address the distinct wrongs and harms of the s. 163.1(4) offence of possessing child pornography, its gravity, and perpetrators’ moral blameworthiness: R. v. Bertrand Marchand, 2023 SCC 26, 487 D.L.R. (4th) 201, at paras. 32-33. This guidance also applies to people who commit the s. 163.1(3) offence of importing child pornography if, like Mr. Scott, they possess the material they import.[^14]

(i) Possession’s Distinct Wrongs and Harms

[144] The extent and harmfulness of child pornography has long been on the rise. In the pre-Internet age, it was more challenging and riskier to produce, distribute, and acquire child pornography through the mail or in-person, and the authorities successfully disrupted many production and distribution networks: Badgley Report, vol. 2, at pp. 1139-1210. The Internet transformed this dynamic by enabling perpetrators to produce, distribute, and acquire child pornography more quickly and easily, and with much less risk of detection. This increased the number