R. v. Mabior, 2012 SCC 47

SUPREME COURT OF CANADA

Date: 20121005 Docket: 33976

Between:

Her Majesty The Queen

Appellant

and

Clato Lual Mabior

Respondent

• and -

Canadian HIV/AIDS Legal Network, HIV & AIDS Legal Clinic Ontario, Coalition des organismes communautaires québécois de lutte contre le sida, Positive Living Society of British Columbia, Canadian AIDS Society, Toronto People With AIDS Foundation, Black Coalition for AIDS Prevention, Canadian Aboriginal AIDS Network, British Columbia Civil Liberties Association, Criminal Lawyers’ Association of Ontario, Association des avocats de la défense de Montréal and Institut national de santé publique du Québec

Interveners

Coram: McLachlin C.J. and LeBel, Deschamps, Fish, Abella, Rothstein, Cromwell, Moldaver and Karakatsanis JJ.

Reasons for Judgment: (paras. 1 to 110)

McLachlin C.J. (LeBel, Deschamps, Fish, Abella, Rothstein, Cromwell, Moldaver and Karakatsanis JJ. concurring)

R. v. Mabior, 2012 SCC 47, [2012] 2 S.C.R. 584

Her Majesty The Queen Appellant

v.

Clato Lual Mabior Respondent

and

Canadian HIV/AIDS Legal Network,

HIV & AIDS Legal Clinic Ontario, Coalition des organismes communautaires québécois de lutte contre le sida, Positive Living Society of British Columbia, Canadian AIDS Society, Toronto People With AIDS Foundation,

Black Coalition for AIDS Prevention, Canadian Aboriginal AIDS Network,

British Columbia Civil Liberties Association,

Criminal Lawyers’ Association of Ontario,

Association des avocats de la défense de Montréal and

Institut national de santé publique du Québec Interveners

Indexed as: R. v. Mabior

2012 SCC 47

File No.: 33976.

2012: February 8; 2012: October 5.

Present: McLachlin C.J. and LeBel, Deschamps, Fish, Abella, Rothstein, Cromwell, Moldaver and Karakatsanis JJ.

on appeal from the court of appeal for manitoba

Criminal law — Aggravated assault — Consent — Fraud — Non‑disclosure of HIV status — Accused undergoing antiretroviral therapy and having protected and unprotected sexual relations knowing he was HIV‑positive — Whether approach outlined in R. v. Cuerrier, 1998 796 (SCC) , [1998] 2 S.C.R. 371 , remains valid in determining whether fraud vitiates consent to sexual relations — Whether non‑disclosure of HIV status in circumstances where no realistic possibility of transmission exists can constitute fraud vitiating consent — Criminal Code, R.S.C. 1985, c. C‑46, ss. 265(3) (c), 268 , 273 .

M was charged with nine counts of aggravated sexual assault based on his failure to disclose his HIV‑positive status to nine complainants before having sex with them ( ss. 265(3) ( c ) and 273 Cr. C. ). None of the complainants contracted HIV. The trial judge convicted him on six of the counts and acquitted him on the other three, on the basis that sexual intercourse using a condom when viral loads are undetectable does not place a sexual partner at “significant risk of serious bodily harm”, as required by Cuerrier . The Court of Appeal varied the decision, holding that either low viral loads or condom use could negate significant risk. This reduced to two the counts on which M could be convicted, and the Court of Appeal entered acquittals on the four remaining counts. The Crown appealed the acquittals.

Held : The appeal should be allowed in part and the convictions in respect of the complaints by S.H., D.C.S. and D.H. should be restored. The appeal should be dismissed in respect of the complaint by K.G.

This Court, in Cuerrier , established that failure to disclose that one has HIV may constitute fraud vitiating consent to sexual relations under s. 265(3) ( c ) Cr. C. Because HIV poses a risk of serious bodily harm, the operative offence is one of aggravated sexual assault ( s. 273 Cr. C. ). To obtain a conviction under ss. 265(3)( c ) and 273, the Crown must show, beyond a reasonable doubt, that the complainant’s consent to sexual intercourse was vitiated by the accused’s fraud as to his HIV status. The test boils down to two elements: (1) a dishonest act (either falsehoods or failure to disclose HIV status); and (2) deprivation (denying the complainant knowledge which would have caused him or her to refuse sexual relations that exposed him or her to a significant risk of serious bodily harm). Failure to disclose may amount to fraud where the complainant would not have consented had he or she known the accused was HIV‑positive, and where sexual contact poses a significant risk of or causes actual serious bodily harm.

Two main criticisms of the Cuerrier test have been advanced: first, that it is uncertain, failing to draw a clear line between criminal and non‑criminal conduct, and second, that it either overextends the criminal law or confines it too closely — the problem of breadth. While it may be difficult to apply, the Cuerrier approach is in principle valid. It carves out an appropriate area for the criminal law — one restricted to “significant risk of serious bodily harm”. The test’s approach to consent accepts the wisdom of the common law that not every deception that leads to sexual intercourse should be criminalized, while still according consent meaningful scope.

The Cuerrier requirement of “significant risk of serious bodily harm” should be read as requiring disclosure of HIV status if there is a realistic possibility of transmission of HIV. This view is supported by the common law and statutory history of fraud vitiating consent to sexual relations, and is in line with Charter values of autonomy and equality that respect the interest of a person to choose whether to consent to sex with a particular person or not. It also gives adequate weight to the nature of the harm involved in HIV transmission, and avoids setting the bar for criminal conviction too high or too low. If there is no realistic possibility of transmission of HIV, failure to disclose that one has HIV will not constitute fraud vitiating consent to sexual relations under s. 265(3)( c ).

The evidence adduced in this case leads to the conclusion that, as a general matter, a realistic possibility of transmission of HIV is negated if: (i) the accused’s viral load at the time of sexual relations was low and (ii) condom protection was used. This general proposition does not preclude the common law from adapting to future advances in treatment and to circumstances where risk factors other than those considered in this case are at play.

Here, the four complainants all consented to sexual intercourse with M, and testified that they would not have had sex with him had they known he was HIV‑positive. M had intercourse by vaginal penetration with the four complainants, during which he ejaculated. At the time of intercourse with the complainants S.H., D.C.S. and D.H., M had a low viral load but did not use a condom. Consequently, those convictions should be maintained. As regards K.G., the record shows that M’s viral load was low. When combined with condom protection, this did not expose K.G. to a significant risk of serious bodily harm. This conviction must accordingly be reversed.

Cases Cited

Applied: R. v. Cuerrier , 1998 796 (SCC) , [1998] 2 S.C.R. 371; referred to: R. v. D.C. , 2012 SCC 48 , [2012] 2 S.C.R. 626; Proprietary Articles Trade Association v. Attorney‑General for Canada , 1931 385 (UK JCPC) , [1931] A.C. 310; Reference re Validity of Section 5(a) of the Dairy Industry Act , 1948 2 (SCC) , [1949] S.C.R. 1, aff’d 1950 342 (UK JCPC) , [1951] A.C. 179; Lord’s Day Alliance of Canada v. Attorney General of British Columbia , 1959 42 (SCC) , [1959] S.C.R. 497; The Queen v. Sault Ste. Marie , 1978 11 (SCC) , [1978] 2 S.C.R. 1299; Boggs v. The Queen , 1981 39 (SCC) , [1981] 1 S.C.R. 49; Skoke‑Graham v. The Queen , 1985 60 (SCC) , [1985] 1 S.C.R. 106; R. v. Roy , 2012 SCC 26 , [2012] 2 S.C.R. 60; R. v. Beatty , 2008 SCC 5 , [2008] 1 S.C.R. 49; The Queen v. Clarence (1888), 22 Q.B.D. 23 ; R. v. Flattery (1877), 13 Cox C.C. 388; R. v. Dee (1884), 15 Cox C.C. 579; R. v. Bennett (1866), 4 F. & F. 1105, 176 E.R. 925 ; R. v. Sinclair (1867), 13 Cox C.C. 28; Hegarty v. Shine (1878) , 14 Cox C.C. 124, aff’d 14 Cox C.C. 145; Papadimitropoulos v. The Queen (1957), 98 C.L.R. 249 ; R. v. Harms (1943), 1943 146 (SK CA) , 81 C.C.C. 4; Bolduc v. The Queen , 1967 57 (SCC) , [1967] S.C.R. 677; R. v. Petrozzi (1987), 1987 2786 (BC CA) , 35 C.C.C. (3d) 528; R. v. Lee (1991), 1991 7351 (ON SC) , 3 O.R. (3d) 726; R. v. Ssenyonga (1993), 1993 14680 (ON SC) , 81 C.C.C. (3d) 257; State v. Marcks , 41 S.W. 973 (1897) , and 43 S.W. 1095 (1898); State v. Lankford , 102 A. 63 (1917); United States v. Johnson , 27 M.J. 798 (1988); United States v. Dumford , 28 M.J. 836 (1989); R. v. Maurantonio , 1967 317 (ON CA) , [1968] 1 O.R. 145; R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45; Application under s. 83.28 of the Criminal Code (Re) , 2004 SCC 42 , [2004] 2 S.C.R. 248; R. v. Ewanchuk , 1999 711 (SCC) , [1999] 1 S.C.R. 330; R. v. B. , [2006] EWCA Crim 2945, [2007] 1 W.L.R. 1567 ; R. v. Mwai , [1995] 3 N.Z.L.R. 149; R. v. Find , 2001 SCC 32 , [2001] 1 S.C.R. 863; R. v. Spence , 2005 SCC 71 , [2005] 3 S.C.R. 458; Twining v. Morrice (1788), 2 Bro. C.C. 326, 29 E.R. 182 ; Conolly v. Parsons (1797), 3 Ves. 625n; Walters v. Morgan (1861), 3 De G. F. & J. 718, 45 E.R. 1056 ; R. v. McCraw , 1991 29 (SCC) , [1991] 3 S.C.R. 72; R. v. Jones , 2002 NBQB 340, [2002] N.B.J. No. 375 (QL) ; R. v. J.A.T. , 2010 BCSC 766 () .

Statutes and Regulations Cited

Act to amend the Criminal Code in relation to sexual offences and other offences against the person and to amend certain other Acts in relation thereto or in consequence thereof , S.C. 1980‑81‑82‑83, c. 125, s. 19.

Canadian Charter of Rights and Freedoms , s. 7.

Crimes Act 1958 (Vic.), ss. 22, 23.

Crimes Act 1961 (N.Z.), 1961, No. 43, ss. 145, 188(2).

Criminal Code , R.S.C. 1985, c. C‑46, ss. 265, 268 , 271(1) , 273 .

Criminal Code, 1892 , S.C. 1892, c. 29, ss. 259( b ), 266.

Criminal Code Act (N.T.), ss. 174C, 174D.

Criminal Law Consolidation Act 1935 (S.A.), s. 29.

Offences against the Person Act, 1861 (U.K.), 24 & 25 Vict., c. 100, ss. 18, 20.

Authors Cited

Bingham, Tom. The Rule of Law . London: Allen Lane, 2010.

Boily, Marie‑Claude, et al. “Heterosexual risk of HIV‑1 infection per sexual act: systematic review and meta‑analysis of observational studies” (2009), 9 Lancet Infect. Dis. 118.

Burris, Scott, et al. “Do Criminal Laws Influence HIV Risk Behavior? An Empirical Trial” (2007), 39 Ariz. St. L.J. 467.

Cohen, Myron S., et al. “Prevention of HIV‑1 Infection with Early Antiretroviral Therapy” (2011), 365 New Eng. J. Med. 493.

Grant, Isabel. “The Prosecution of Non‑disclosure of HIV in Canada: Time to Rethink Cuerrier ” (2011), 5 M.J.L.H. 7.

Leigh, L. H. “Two cases on consent in rape” (2007), 5 Arch. News 6.

Nightingale, Brenda L. The Law of Fraud and Related Offences . Scarborough, Ont.: Carswell, 1996 (loose‑leaf updated 2011, release 3).

Wainberg, Mark A. “Criminalizing HIV transmission may be a mistake” (2009), 180 C.M.A.J. 688.

Weller, Susan C., and Karen Davis‑Beaty. “Condom effectiveness in reducing heterosexual HIV transmission” (2002), 1 Cochrane Database Syst. Rev. CD003255.

APPEAL from a judgment of the Manitoba Court of Appeal (Steel, MacInnes and Beard JJ.A.), 2010 MBCA 93 , 258 Man. R. (2d) 166, 261 C.C.C. (3d) 520, 79 C.R. (6th) 1, [2011] 2 W.W.R. 211, [2010] M.J. No. 308 (QL), 2010 CarswellMan 587, setting aside convictions for aggravated sexual assault entered by McKelvey J., 2008 MBQB 201 , 230 Man. R. (2d) 184, [2008] M.J. No. 277 (QL), 2008 CarswellMan 406. Appeal allowed in part.

Elizabeth Thomson and Ami Kotler , for the appellant.

Amanda Sansregret and Corey La Berge , for the respondent.

Jonathan Shime , Corie Langdon , Richard Elliott and Ryan Peck , for the interveners the Canadian HIV/AIDS Legal Network, the HIV & AIDS Legal Clinic Ontario, Coalition des organismes communautaires québécois de lutte contre le sida, the Positive Living Society of British Columbia, the Canadian AIDS Society, the Toronto People With AIDS Foundation, the Black Coalition for AIDS Prevention and the Canadian Aboriginal AIDS Network.

Michael A. Feder and Angela M. Juba , for the intervener the British Columbia Civil Liberties Association.

P. Andras Schreck and Candice Suter , for the intervener the Criminal Lawyers’ Association of Ontario.

François Dadour , for the intervener Association des avocats de la défense de Montréal.

Lucie Joncas and François Côté , for the intervener Institut national de santé publique du Québec.

The judgment of the Court was delivered by

The Chief Justice —

I. Overview

[ 1 ] This case raises the issue of whether an HIV-positive person who engages in sexual relations without disclosing his condition commits aggravated sexual assault.

[ 2 ] Sex without consent is sexual assault under s. 265 of the Criminal Code , R.S.C. 1985, c. C-46. R. v. Cuerrier , 1998 796 (SCC) , [1998] 2 S.C.R. 371, establishes that failure to advise a partner of one’s HIV status may constitute fraud vitiating consent. Because HIV poses a risk of serious bodily harm, the operative offence is one of aggravated sexual assault, attracting a maximum sentence of life imprisonment: Cuerrier , at para. 95 ; ss. 265 , 268 and 273 Cr. C.

[ 3 ] While Cuerrier laid down the basic requirements for the offence, the precise circumstances when failure to disclose HIV status vitiates consent and converts sexual activity into a criminal act remain unclear. The parties ask this Court for clarification.

[ 4 ] I conclude that a person may be found guilty of aggravated sexual assault under s. 273 of the Criminal Code if he fails to disclose HIV-positive status before intercourse and there is a realistic possibility that HIV will be transmitted. If the HIV-positive person has a low viral count as a result of treatment and there is condom protection, the threshold of a realistic possibility of transmission is not met, on the evidence before us.

II. Background

[ 5 ] The respondent, Mr. Mabior, lived in Winnipeg. His house was a party place. People came in and out, including a variety of young women. Alcohol and drugs were freely dispensed. From time to time, Mr. Mabior had sex with women who came to his house, including the nine complainants in this case.

[ 6 ] Mr. Mabior did not tell the complainants that he was HIV-positive before having sex with them; indeed, he told one of them that he had no STDs. On some occasions, he wore condoms, on others he did not. Sometimes the condoms broke or were removed, and in some cases, the precise nature of the protections taken is unclear. Eight of the nine complainants testified that they would not have consented to sex with Mr. Mabior had they known he was HIV-positive. None of the complainants contracted HIV.

[ 7 ] Mr. Mabior was charged with nine counts of aggravated sexual assault (and other related offences), based on his failure to disclose to the complainants that he was HIV-positive. In defence, Mr. Mabior called evidence that he was under treatment, and that he was not infectious or presented only a low risk of infection at the relevant times.

[ 8 ] The trial judge convicted Mr. Mabior of six counts of aggravated sexual assault ( 2008 MBQB 201 , 230 Man. R. (2d) 184). She acquitted him on the other three, on the basis that sexual intercourse using a condom when viral loads are undetectable does not place a sexual partner at “significant risk of serious bodily harm”, as required by Cuerrier .

[ 9 ] Mr. Mabior appealed from these six convictions; the Crown did not appeal from the three acquittals. The Manitoba Court of Appeal varied the trial judge’s decision, holding that either low viral loads or condom use could negate significant risk ( 2010 MBCA 93 , 258 Man. R. (2d) 166). This reduced to two the counts on which Mr. Mabior could be convicted, and the Court of Appeal entered acquittals on the four remaining counts. The Crown appeals these acquittals. Mr. Mabior has not cross-appealed against the two convictions upheld by the Court of Appeal.

III. The Legislation

[ 10 ] Sections 265 and 273 of the Criminal Code provide:

265. (1) A person commits an assault when

( a ) without the consent of another person, he applies force intentionally to that other person, directly or indirectly;

(2) This section applies to all forms of assault, including sexual assault . . . and aggravated sexual assault.

(3) For the purposes of this section, no consent is obtained where the complainant submits or does not resist by reason of

( c ) fraud ;

273. (1) Every one commits an aggravated sexual assault who, in committing a sexual assault, wounds, maims, disfigures or endangers the life of the complainant .

(2) Every person who commits an aggravated sexual assault is guilty of an indictable offence and liable

( b ) in any other case, to imprisonment for life .

IV. The Issues

[ 11 ] The issues are as follows:

A. What is the correct interpretation of “fraud” vitiating consent to sexual activity in s. 265(3) ( c ) of the Criminal Code ?

(1) Problems with the existing interpretation of “fraud” vitiating consent:

(a) Uncertainty;

(b) Breadth.

(2) Guides to interpretation:

(a) The purposes of the criminal law;

(b) The common law and statutory history of fraud vitiating consent to sexual relations;

(c) Charter values;

(d) The experience of other common law jurisdictions.

(3) Finding a solution:

(a) The active misrepresentation approach;

(b) The absolute disclosure approach;

(c) A case-by-case fact-based approach;

(d) Judicial notice;

(e) Relationship-based distinctions;

(f) The reasonable partner approach;

(g) An evolving common law approach.

(4) Realistic possibility of HIV transmission.

V. Discussion

A. What is the Correct Interpretation of “Fraud” Vitiating Consent to Sexual Activity in Section 265(3) (c) of the Criminal Code ?

(1) Problems With the Existing Interpretation of “Fraud” Vitiating Consent

[ 12 ] This Court considered “fraud” under s. 265(3) ( c ) 14 years ago in Cuerrier . The majority test in Cuerrier may be stated in different ways, but boils down to two elements: (1) a dishonest act (either falsehoods or failure to disclose HIV status); and (2) deprivation (denying the complainant knowledge which would have caused her to refuse sexual relations that exposed her to a significant risk of serious bodily harm).

[ 13 ] We are invited to revisit the Cuerrier test by the parties and by the Manitoba Court of Appeal in this case and the Quebec Court of Appeal in the companion case of R. v. D.C. , 2012 SCC 48 , [2012] 2 S.C.R. 626. Two main criticisms of the Cuerrier test are advanced: first, that it is uncertain, failing to draw a clear line between criminal and non-criminal conduct (uncertainty), and second, that it either overextends the criminal law or confines it too closely — the problem of breadth.

[ 14 ] I turn first to the criticism that the Cuerrier test is uncertain. It is a fundamental requirement of the rule of law that a person should be able to predict whether a particular act constitutes a crime at the time he commits the act. The rule of law requires that laws provide in advance what can and cannot be done: Lord Bingham, The Rule of Law (2010). Condemning people for conduct that they could not have reasonably known was criminal is Kafkaesque and anathema to our notions of justice. After-the-fact condemnation violates the concept of liberty in s. 7 of the Canadian Charter of Rights and Freedoms and has no place in the Canadian legal system.

[ 15 ] The Cuerrier test gives rise to two uncertainties — what constitutes “significant risk” and what constitutes “serious bodily harm”? These terms are broad and different people can and do read them in different ways.

[ 16 ] About “significant risk”, some people say that virtually any risk of serious bodily harm is significant. Others argue that to be significant, the risk must rise to a higher level. These debates centre on statistical percentages. Is a 1% risk “significant”? Or should it be 10% or 51% or, indeed, .01%? How is a prosecutor to know or a judge decide? And if prosecutors, defence counsel and judges debate the point, how — one may ask — is the ordinary Canadian citizen to know? This uncertainty is compounded by the fact that a host of variables may affect the actual risk of infection.

[ 17 ] Debate has also surrounded the requirement that the risk be one of “serious bodily harm”. Some sexually transmitted diseases (“STDs”) involve little beyond treatable temporary discomfort. Yet even that discomfort, while it persists, may be serious from the perspective of the victim. Other STDs, like HIV, are extremely serious, involving permanent and life-altering symptoms, and in some cases death. Between these two extremes lie many other STDs, some more debilitating than others. Which are sufficiently serious to attract the sanction of the criminal law? Cuerrier offers no clear answer.

[ 18 ] The uncertainty inherent in the concepts of significant risk and serious bodily harm is compounded by the fact that they are interrelated. The more serious the nature of the harm, the lower the probability of transmission need be to amount to a significant risk of serious bodily harm, it is argued. So it is not simply a matter of percentage of risk and seriousness of the potential disease. It is a matter of the two as they relate to each other.

[ 19 ] What emerges is a complex calculus that makes it impossible, in many cases, to predict in advance whether a particular act is criminal under s. 265(3) ( c ) or not. The second major criticism of Cuerrier relates to the scope of the conduct it catches. The danger of an overbroad interpretation is the criminalization of conduct that does not present the level of moral culpability and potential harm to others appropriate to the ultimate sanction of the criminal law. A criminal conviction and imprisonment, with the attendant stigma that attaches, is the most serious sanction the law can impose on a person, and is generally reserved for conduct that is highly culpable — conduct that is viewed as harmful to society, reprehensible and unacceptable. It requires both a culpable act — actus reus — and a guilty mind — mens rea — the parameters of which should be clearly delineated by the law.

(2) Guides to Interpretation

[ 20 ] As for all issues of statutory interpretation, the basic question is what Parliament intended. That intention is discovered by looking at the words of the provision, informed by its history, context and purpose.

[ 21 ] The words of s. 265(3)( c ) do not on their face reveal much about how Parliament intended “fraud” to be interpreted. The concept was taken by Parliament from the common law.

[ 22 ] The interpretation of “fraud” vitiating consent to sexual relations is informed by four considerations: (a) the purposes of the criminal law; (b) the common law and statutory history of the concept; (c) Charter values, particularly equality, autonomy, liberty, privacy and human dignity; and (d) the experience of other common law jurisdictions. I will consider each in turn.

(a) The Purposes of the Criminal Law

[ 23 ] The interpretation of fraud vitiating consent to sexual relations should further the purposes of the criminal law, notably identifying, deterring and punishing criminal conduct, defined by a wrongful act and guilty mind. Morality infuses the criminal law. But the law does not seek to criminalize all immorality. The principal objective of the criminal law is the public identification of wrongdoing qua wrongdoing which violates public order and is so blameworthy that it deserves penal sanction: Proprietary Articles Trade Association v. Attorney-General for Canada , 1931 385 (UK JCPC) , [1931] A.C. 310 (P.C.); Reference re Validity of Section 5(a) of the Dairy Industry Act , 1948 2 (SCC) , [1949] S.C.R. 1, aff’d 1950 342 (UK JCPC) , [1951] A.C. 179 (P.C.); Lord’s Day Alliance of Canada v. Attorney General of British Columbia , 1959 42 (SCC) , [1959] S.C.R. 497; The Queen v. Sault Ste. Marie , 1978 11 (SCC) , [1978] 2 S.C.R. 1299; Boggs v. The Queen , 1981 39 (SCC) , [1981] 1 S.C.R. 49; Skoke-Graham v. The Queen , 1985 60 (SCC) , [1985] 1 S.C.R. 106.

[ 24 ] The law draws a sharp distinction between civil wrongdoing and criminal wrongdoing. Criminal conduct requires both a wrongful act and a guilty mind. It requires “a significant fault element”: R. v. Roy , 2012 SCC 26 , [2012] 2 S.C.R. 60, at para. 32 . As Charron J. stated for the majority of this Court in R. v. Beatty , 2008 SCC 5 , [2008] 1 S.C.R. 49, at para. 34 :

If every departure from the civil norm is to be criminalized, regardless of the degree, we risk casting the net too widely and branding as criminals persons who are in reality not morally blameworthy. Such an approach risks violating the principle of fundamental justice that the morally innocent not be deprived of liberty.

The potential consequences of a conviction for aggravated sexual assault — up to life imprisonment — underline the importance of insisting on moral blameworthiness in the interpretation of s. 265(3) ( c ) of the Criminal Code .

(b) The Common Law and Statutory History of Fraud Vitiating Consent to Sexual Relations

[ 25 ] The common law history of fraud vitiating consent to sexual relations reveals three periods. The early cases support the view that failure to disclose to a partner the fact that one has a serious sexually transmitted disease could constitute fraud vitiating consent to sexual relations, resulting in convictions for rape or assault. This was reversed in The Queen v. Clarence (1888), 22 Q.B.D. 23 (Cr. Cas. Res.) , which held that fraud was confined to deception as to the sexual nature of the act or as to the identity of the male sexual partner. In the post- Charter era, a return to a broader view of fraud vitiating consent is appropriate.

[ 26 ] The first recorded cases to look at the problem before us took a generous approach to consent to sexual intercourse — one that accepted that sexual partners (always women in those days) were entitled to refuse sexual intercourse and should not be tricked into it by deceit. The courts adopted a flexible approach to “fraud” vitiating consent to sexual relations. Without attempting to define the term, they showed themselves willing to extend the term to fundamental aspects of sexual intercourse.

[ 27 ] For example, in R. v. Flattery (1877), 13 Cox C.C. 388 (C.C.A.), a conviction of rape was upheld for a man operating a booth at a fair who had obtained sex from a girl of 19 on the pretext of medical treatment. It was held that the victim’s consent to physical contact with the accused was vitiated by his fraud, because she had only consented to a surgical operation, and not to a sexual act.

[ 28 ] Similarly, early common law cases accepted that impersonation of the spouse — falsely pretending to be the victim’s husband — could constitute fraud vitiating consent. In R. v. Dee (1884), 15 Cox C.C. 579 (Cr. Cas. Res. Ir.), O’Brien J. left no doubt on this issue:

That brings us back to the question which in law is the crime of rape. The crime is the invasion of a woman’s person without her consent, and I see no real difference between the want of consent and the act being against her will, which is the language of the indictment, though the distinction is taken by Lord Campbell, or between the negation of consent and positive dissent. Whether the act of consent be the result of overpowering force, or of fear, or of incapacity, or of natural condition, or of deception, it is still want of consent, and the consent must be, not consent to the act, but to the act of the particular person — not in the abstract, but in the concrete . . . . [p. 598]

[ 29 ] In R. v. Bennett (1866), 4 F. & F. 1105, 176 E.R. 925 (West. Cir.) , similar reasoning was applied to hold that concealment of venereal disease amounted to fraud vitiating consent:

An assault is within the rule that fraud vitiates consent, and therefore, if the prisoner, knowing that he had a foul disease, induced his niece to sleep with him, intending to possess her, and infected her, she being ignorant of his condition, any consent, which she may have given, would be vitiated, and the prisoner would be guilty of an indecent assault. [p. 925]

[ 30 ] Again in R. v. Sinclair (1867), 13 Cox C.C. 28, the Central Criminal Court found fraud vitiating consent for non-disclosure of gonorrhœa, stating that if the complainant “would not have consented if she had known the fact, then her consent is vitiated by the deceit practised upon her, and the prisoner would be guilty of an assault” (p. 29).

[ 31 ] These cases evinced a generous approach to the issue of consent and when deceit might vitiate it, an approach that respected the right of the women involved to choose whether to have intercourse or not. However, this jurisprudence was shortly to be set aside, in a series of cases which culminated in Clarence . To read these cases is to enter a world foreign to modern sensitivities — the world of Victorian morality.

[ 32 ] The case that announced the change was Hegarty v. Shine (1878), 14 Cox C.C. 124 (H.C.J. Ir. (Q.B.D.)), a civil case involving an action for assault. Mr. Shine, the master of the house, had sexual relations with his domestic servant over a period of two years. She became pregnant and had a child. Both she and the child were infected with syphilis. The court dismissed the action against Mr. Shine on the basis of ex turpi causa non oritur actio — the plaintiff was the victim of her own immoral act, which the law could not condone. On the matter of fraud, the majority stated that the doctrine of fraud was confined to mistake as to whether the act was sexual or not : “In the case before us the defendant actively consented to the very thing, that is to say, sexual intercourse, with full knowledge and experience of the nature of the act” (p. 130). The appeal court ((1878), 14 Cox C.C. 145) confirmed that only deceit as to the nature of the act could vitiate consent. The court emphasized the pitiable condition of the victim, but concluded the law could not assist her.

[ 33 ] The point of no return for the earlier, more open view of fraud was reached in Clarence . That case confirmed that fraud could vitiate consent to sexual relations only if the complainant was deceived as to the sexual nature of the act or as to the identity of the man. The facts were simple. The couple were married. The husband did not tell his wife he had gonorrhœa and infected her. The husband was charged with assault and unlawful infliction of bodily harm.

[ 34 ] The fact that 13 judges sat suggests the case was viewed as important. The court divided nine to four, and the husband was acquitted. The majority held that fraud in the context of sexual relations had been interpreted too broadly by the earlier cases, and that it was necessary to limit its application to situations where the complainant was deceived as to the sexual nature of the act or as to the identity of the man. This produced a rule that was to prevail for almost 100 years that fraud could not vitiate consent to sexual intercourse unless it went to the “sexual nature of the act” or to the identity of the sexual partner.

[ 35 ] The opinion of Stephen J. encapsulates the view of the majority in Clarence . Stephen J. found that the offence of unlawful infliction of bodily harm could not apply to the case. While the act posed by the husband was indeed unlawful — infecting one’s wife was forbidden by the law relating to marriage since it constituted cruelty and could be evidence of adultery —, it could not be said that it constituted infliction of bodily harm. Indeed, the words “infliction of bodily harm” were construed as requiring a physical assault. Stephen J. found that infecting a person with a disease did not constitute such an assault.

[ 36 ] Stephen J. went on to consider the question of obtaining sexual relations by fraud. He expressed the view that the only fraud capable of vitiating consent to sexual relations was fraud as to the nature of the act of intercourse, or as to the identity of the sexual partner. If the victim knew the act was sexual, and was not deceived as to the identity of her partner, she could not complain that she had been deceived and her “consent” fraudulently obtained. Stephen J. commented briefly that neither Bennett nor Sinclair could be relied upon as precedent.

[ 37 ] The reasons of Pollock B., also in the majority, added that sexual acts done by a husband to his wife cannot be unlawful (pp. 63-64). The husband possessed conjugal rights over his wife to which she had consented by marrying him. Once married, the wife had no right to refuse her husband’s demands. Since sexual acts between spouses were lawful, Pollock B. reasoned, all such acts done by a husband — including those characterized by cruelty — must be lawful.

[ 38 ] The Clarence test was accepted throughout the common law world and prevailed until recent times. The strictness with which it was applied is illustrated by a 1957 decision of the Australian High Court, Papadimitropoulos v. The Queen (1957), 98 C.L.R. 249 . The accused had induced the complainant to have sexual intercourse by duping her into believing that they were legally married. He was acquitted of the charge of rape. The court summed up the law as follows:

To say that in having intercourse with him she supposed that she was concerned in a perfectly moral act is not to say that the intercourse was without her consent. To return to the central point; rape is carnal knowledge of a woman without her consent: carnal knowledge is the physical fact of penetration; it is the consent to that which is in question; such a consent demands a perception as to what is about to take place, as to the identity of the man and the character of what he is doing. But once the consent is comprehending and actual the inducing causes cannot destroy its reality and leave the man guilty of rape. [Emphasis added; p. 261.]

[ 39 ] The views of the majority in Clarence were reflected in the first Canadian Criminal Code in 1892 (S.C. 1892, c. 29). Parliament defined fraud for purposes of rape and indecent assault narrowly, by restricting it to “false and fraudulent representations as to the nature and quality of the act”: ss. 259( b ) and 266. The Code thus incorporated the concerns of the majority in Clarence . The deceitful act was limited to “false and fraudulent representations” by opposition to simple concealment, and the subject of the fraud was limited to the “nature and quality of the act”. As a consequence, Canadian courts accepted Clarence as settled law and continued to hold that only active fraud as to the nature of the act, i.e. fraud as to its sexual character, or as to the identity of one’s sexual partner constituted fraud vitiating consent to sexual intercourse: see, e.g., R. v. Harms (1943), 1943 146 (SK CA) , 81 C.C.C. 4 (Sask. C.A.); Bolduc v. The Queen , 1967 57 (SCC) , [1967] S.C.R. 677.

[ 40 ] In 1983, Parliament amended the Criminal Code to create the present s. 265(3) ( c ): S.C. 1980-81-82-83, c. 125, s. 19. This amendment was part of a major overhaul of the law of sexual offences which aimed, inter alia , at protecting the integrity of the person and eliminating sexual discrimination in the criminal law. The new provision referred simply to “fraud”, dropping the qualifying phrases “false and fraudulent representations” and “nature and quality of the act”. Arguably, this change evidenced Parliament’s intent that “fraud” should be read more broadly than it had been in the past. However, courts continued to apply a restrictive interpretation to the term: see R. v. Petrozzi (1987), 1987 2786 (BC CA) , 35 C.C.C. (3d) 528 (B.C.C.A.); R. v. Lee (1991), 1991 7351 (ON SC) , 3 O.R. (3d) 726 (Gen. Div.); R. v. Ssenyonga (1993), 1993 14680 (ON SC) , 81 C.C.C. (3d) 257 (Ont. Ct. (Gen. Div.)).

[ 41 ] Occasionally, a more generous approach surfaced as to when fraud could vitiate consent to sexual relations. For instance, American caselaw was more inclined to convict the accused of assault or rape on facts similar to those in Clarence : State v. Marcks , 41 S.W. 973 (Miss. 1897), at p. 973, and 43 S.W. 1095 (1898), at pp. 1097-98; State v. Lankford , 102 A. 63 (Del. Ct. Gen. Sess. 1917), at p. 64; United States v. Johnson , 27 M.J. 798 (A.F.C.M.R. 1988), at p. 804; United States v. Dumford , 28 M.J. 836 (A.F.C.M.R. 1989), at p. 839. And in Canada, in R. v. Maurantonio , 1967 317 (ON CA) , [1968] 1 O.R. 145 (C.A.), Hartt J. ( ad hoc ) held that “the words ‘nature and quality of the act’ . . . should not be so narrowly construed as to include only the physical action but rather must be interpreted to encompass those concomitant circumstances which give meaning to the particular physical activity in question” (p. 153). But these were minority voices.

[ 42 ] There are not many cases on the books, to be sure. One understands why when one considers the practical implications of the Clarence test. The cases where a woman consents to sex but thinks it is not sex or that it is sex with a different man are necessarily rare. Submitting to a medical examination only to discover it is a sexual encounter might be such a case; the complainant does not consent to the sexual act but to a medical procedure. Under Clarence and its progeny, if the woman consented to the sexual act with the given man, no matter what the deceit, the man could not be convicted for his act. It is not surprising that the legal lexicon contains few cases dealing with fraud vitiating consent. The stark fact was that in all but rare cases fraud could not vitiate consent.

[ 43 ] Canadian common law on fraud vitiating consent to sexual relations has now entered a third, post- Clarence era. Charter values of equality, autonomy, liberty, privacy and human dignity require full recognition of the right to consent or to withhold consent to sexual relations. Fraud under s. 265(3) ( c ) must be interpreted with these values in mind. The Clarence line of jurisprudence, which confined fraud to the question of whether the complainant knew the act was sexual or not, is no longer appropriate in the Canadian context. To hold that a complainant consents to the risk of an undisclosed serious disease because he or she knew the act was sexual affronts contemporary sensibilities and contemporary constitutional values.

(c) Charter Values

[ 44 ] Courts must interpret legislation harmoniously with the constitutional norms enshrined in the Charter : R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45, at para. 33 ; Application under s. 83.28 of the Criminal Code (Re) , 2004 SCC 42 , [2004] 2 S.C.R. 248, at para. 35 . Charter values are always relevant to the interpretation of a disputed provision of the Criminal Code .

[ 45 ] The Charter values of equality, autonomy, liberty, privacy and human dignity are particularly relevant to the interpretation of fraud vitiating consent to sexual relations. The formerly narrow view of consent has been replaced by a view that respects each sexual partner as an autonomous, equal and free person. Our modern understanding of sexual assault is based on the preservation of the right to refuse sexual intercourse: sexual assault is wrong because it denies the victim’s dignity as a human being. Fraud in s. 265(3) ( c ) of the Criminal Code must be interpreted in light of these values.

[ 46 ] As we have already seen, prior to the adoption of the Charter in 1982 and the reform of sexual offences in 1983, courts took a restrictive view of how lack of consent to sexual relations could be established and how consent could be negated by fraud. Rules of evidence and procedure, like the ancient rule that non-consent must be supported by evidence of a “hue and cry” in the neighbourhood immediately after the alleged sexual assault, or the willingness of judges to infer consent from dress or prior sexual experience, systemically biased the trial process in favour of finding consent . In like fashion, the jurisprudence, post- Clarence , took a narrow view of fraud capable of vitiating consent, holding that it went only to the sexual nature of the act, and that it did not apply to married women, who were bound to submit to their husbands in all circumstances.

[ 47 ] Post- Charter Canadian law has repudiated this crabbed view of consent and fraud. Amendments to the Criminal Code have removed the evidentiary burdens and presumptions that once made proof of lack of consent difficult. Courts have held that judges may not infer consent from the way the complainant was dressed or the fact that she may have flirted: R. v. Ewanchuk , 1999 711 (SCC) , [1999] 1 S.C.R. 330 . And in 1998, Cuerrier signaled a return to a generous interpretation of fraud capable of vitiating consent.

[ 48 ] In keeping with the Charter values of equality and autonomy, we now see sexual assault not only as a crime associated with emotional and physical harm to the victim, but as the wrongful exploitation of another human being. To engage in sexual acts without the consent of another person is to treat him or her as an object and negate his or her human dignity. Although the Charter is not directly engaged, the values that animate it must be taken into account in interpreting s. 265(3) ( c ) of the Criminal Code .

(d) The Experience of Other Common Law Jurisdictions

[ 49 ] The parties and a number of interveners have relied on the approach of foreign legal systems to support their positions.

[ 50 ] A survey of comparative law shows that common law jurisdictions criminalize the actual sexual transmission of HIV, when the HIV-positive person is aware of his or her serologic status and when the partner does not give informed consent to the risk of infection. Several jurisdictions treat transmission of HIV following non-disclosure as offences involving bodily harm, rather than as sexual offences. Such transmission has been prosecuted on the following legal grounds: in England, for reckless infliction of bodily injury, under s. 20 of The Offences against the Person Act, 1861 , 24 & 25 Vict., c. 100; in the Australian state of Victoria, for “[c]onduct endangering life” (s. 22 of the Crimes Act 1958 (Vic.)); and in New Zealand, for, inter alia , “reckless disregard for the safety of others” that “causes grievous bodily harm” (s. 188(2) of the Crimes Act 1961 , No. 43). See I. Grant, “The Prosecution of Non-disclosure of HIV in Canada: Time to Rethink Cuerrier ” (2011), 5 M.J.L.H. 7, at pp. 31-41.

[ 51 ] However, deception that exposes a partner to a risk of transmission — but that does not ultimately result in transmission — is not criminalized in many jurisdictions. In England, exposing one’s partner to HIV without advising of HIV status does not vitiate consent. Fraud vitiates consent only if it goes to “the nature or purpose of the relevant act”, which excludes deception as to HIV status: R. v. B. , [2006] EWCA Crim 2945, [2007] 1 W.L.R. 1567, at para. 12 . Theoretically, s. 18 of The Offences against the Person Act, 1861 , which criminalizes intentional infliction of harm, could be used in England to punish non-disclosure of HIV-positive status as a form of attempted intentional infliction of harm. However, the high evidentiary burden “of proving that someone actually intended to transmit HIV through sex (as opposed to intending to have unprotected sex)” places a prohibitive onus on the Crown (Grant, at p. 32).

[ 52 ] In Australia, six of the nine jurisdictions do not criminalize exposure absent transmission. Of the three jurisdictions that criminalize exposure to a risk of transmission ( Crimes Act 1958 (Vic.), ss. 22 and 23; Criminal Law Consolidation Act 1935 (S.A.), s. 29; Criminal Code Act (N.T.), ss. 174C and 174D), two (Victoria and the Northern Territory) apply lesser criminal offences for cases where no actual transmission resulted from exposure.

[ 53 ] In New Zealand, liability hinges, inter alia , on whether transmission occurs. Absent transmission, the failure to disclose HIV-positive status prior to sexual relations attracts the lesser offence of criminal nuisance (s. 145 of the Crimes Act 1961 ; R. v. Mwai , [1995] 3 N.Z.L.R. 149 (C.A.)).

[ 54 ] Professor Grant summarizes the contrast between the Canadian approach and the approach taken by England, Australia and New Zealand as follows:

In Canada, the same charge of aggravated (sexual) assault is typically used regardless of the nature of the deception, whether the virus is transmitted, or whether there is an isolated incident of non-disclosure or an ongoing course of non-disclosure. . . . In all other jurisdictions [canvassed in this study], the offence is characterized as the infliction of bodily harm, and not as non-consensual sexual contact. [p. 42]

[ 55 ] In sum, while the experience of other jurisdictions is not conclusive, it sounds a note of caution against extending the criminal law beyond its appropriate reach in this complex and emerging area of the law.

(3) Finding a Solution

[ 56 ] We have discussed the need for a clear and appropriately tailored test for fraud vitiating consent in s. 265(3) ( c ) of the Criminal Code , in the context of failure to disclose HIV-positive status. With a view to meeting that need, we have canvassed four guides to construing the provision — the proper ambit of the criminal law, the common law and statutory background of the concept of fraud, Charter values and the approach to non-disclosure of HIV status adopted in other countries. This brings us to the nub of the question before us — when, precisely, should non-disclosure of HIV status amount to fraud vitiating consent under s. 265(3) ( c )?

[ 57 ] The four interpretational considerations suggest that the Cuerrier test is valid from the perspective of principle and impact. The Cuerrier test, to recap, requires proof of two elements: (1) a dishonest act; and (2) deprivation. It defines the first element broadly in terms of either misrepresentation or non-disclosure of HIV, and the second element equally broadly in terms of whether the act poses a “significant risk of serious bodily harm”.

[ 58 ] While it may be difficult to apply, the Cuerrier approach is in principle valid. It carves out an appropriate area for the criminal law — one restricted to “significant risk of serious bodily harm”. It reflects the Charter values of autonomy, liberty and equality, and the evolution of the common law, appropriately excluding the Clarence line of authority. The test’s approach to consent accepts the wisdom of the common law that not every deception that leads to sexual intercourse should be criminalized, while still according consent meaningful scope. While Cuerrier takes the criminal law further than courts in other common law jurisdictions have, it can be argued other courts have not gone far enough: see L. H. Leigh, “Two cases on consent in rape” (2007), 5 Arch. News 6.

[ 59 ] Some interveners challenge the use of the criminal law in the case of HIV on the ground that it may deter people from seeking treatment or disclosing their condition, thereby increasing the health risk to the carrier and those he has sex with. On the record before us, I cannot accept this argument. The only “evidence” was studies presented by interveners suggesting that criminalization “probably” acts as a deterrent to HIV testing: see, e.g., M. A. Wainberg, “Criminalizing HIV transmission may be a mistake” (2009), 180 C.M.A.J. 688. Other studies suggest little difference in reporting rates in states that criminalized and did not criminalize behaviour: S. Burris, et al., “Do Criminal Laws Influence HIV Risk Behavior? An Empirical Trial” (2007), 39 Ariz. St. L.J. 467, at p. 501. The conclusions in these studies are tentative, and the studies were not placed in evidence and not tested by cross-examination. They fail to provide an adequate basis to justify judicial reversal of the accepted place of the criminal law in this domain.

[ 60 ] It follows that Cuerrier should not be jettisoned. The problems of uncertainty and appropriate reach that have emerged in its application should be addressed, but to the extent possible within the general framework of Cuerrier . This brings us to the suggestions for dealing with the problems of uncertainty and overbreadth that have arisen in applying Cuerrier .

(a) The Active Misrepresentation Approach

[ 61 ] It is argued that one way to correct the uncertainty and reach problems in Cuerrier ’s application is to narrow the concept of dishonest act to active deception — a clear misrepresentation or a lie in response to a clear question.

[ 62 ] This approach would go a long way to remedying the problem of uncertainty. A person would know that unless they made a positive misrepresentation or lied in response to a clear question, they would not be committing a crime. And it would also remedy the perceived problem of overbreadth; the narrower definition of fraud would prevent those who may inadvertently or negligently fail to disclose HIV status from being criminalized.

[ 63 ] This approach has some support. My concurring reasons in Cuerrier defined fraud as arising when “the person represents that he or she is disease-free” (para. 72, Gonthier J. concurring). This “active deception” approach has been viewed favourably by some: see Leigh, at p. 7.

[ 64 ] The approach, however, also suffers from difficulties. The first is that it may be difficult to distinguish between active deception and passive deception by non-disclosure. Everything depends on the circumstances. This approach’s sensitivity to context blurs the line between criminal and non-criminal conduct. Conversations leading up to sexual acts are not always models of clarity. People tend not to use precise language. Gestures may supplant language. For example, is the shake of the head in response to a partner’s question about HIV a positive representation that one does not have HIV? Moreover, there may be circumstances where the partner may reasonably infer absence of disease from prior conversations or circumstances. To be sure, trial judges can untangle complicated fact situations and decide if there has been a positive misrepresentation. But the fact remains that this marker may not fully or fairly capture the deception that underlies a particular fraud, and may be a difficult test to apply in practice.

[ 65 ] Second, there is arguably no principled distinction between active deception and passive deception. Should the trusting wife who does not ask a direct question as to HIV status of her partner be placed in a worse position than the casual date who does? The point from the complainant’s perspective is that in either case she would not have given her consent had the deception not occurred. Fraud is fraud, whether induced by blatant lies or sly deceit.

(b) The Absolute Disclosure Approach

[ 66 ] The Crown submits that all HIV-positive people should be required to disclose the fact to all sexual partners in all cases. This amounts to removing the “significant risk of serious bodily harm” requirement from the second element of the Cuerrier test.

[ 67 ] This approach provides a clear test, solving the problem of uncertainty. However, its solution to the problem of breadth is less convincing. This approach arguably casts the net of criminal culpability too widely. People who act responsibly and whose conduct causes no harm and indeed may pose no risk of harm, could find themselves criminalized and imprisoned for lengthy periods. Moreover, this approach seems to expand fraud vitiating consent in s. 265(3) ( c ) further than necessary, by defining it as simple dishonesty and effectively eliminating the deprivation element of fraud. Finally, this absolute approach is arguably unfair and stigmatizing to people with HIV, an already vulnerable group. Provided people so afflicted act responsibly and pose no risk of harm to others, they should not be put to the choice of disclosing their disease or facing criminalization.

(c) A Case-by-Case Fact-based Approach

[ 68 ] The respondent argues that the Cuerrier “significant risk of serious bodily harm” test remains a workable test, but seeks to clarify it by emphasizing that it is simply a statement of what the evidence must establish to support a finding of fraud under s. 265(3) ( c ). A “significant risk of serious bodily harm” must be established by medical evidence in each case. The question is whether at the time of the sexual act in question, the particular act posed a significant risk of transmitting HIV. This typically requires the Crown to call expert evidence as to the accused’s viral count at the time of the offence as well as risks associated with any condom protection used. The trial judge must be satisfied that the evidence establishes beyond a reasonable doubt that the accused’s conduct at the moment of the offence posed a significant risk of serious bodily harm. Any doubt must be resolved in favour of the accused.

[ 69 ] While respectful of the rights of accused persons, this approach does not remedy the problems of uncertainty and reach that make Cuerrier difficult to apply. The process would be onerous. In every case, medical experts would have to be called. Lengthy examination and cross-examination would have to take place. Trial judges would have to spend long hours assessing the evidence to determine if it establishes a “significant risk of serious bodily harm” at the time of the alleged offence. Finally, the risk of conflicting judgments could render the process unfair from a systemic standpoint. The court of appeal, while accepting the trial judge’s conclusions on the evidence, might take a different view on the mixed question of fact and law of whether the risk was “significant”. Years may pass in legal no-man’s-land with no one knowing whether the accused is guilty or not guilty. Enormous costs, both for the prosecution and the defence, would be run up. This case illustrates all these problems.

(d) Judicial Notice of the Effect of Condom Use

[ 70 ] To mitigate the uncertainties of the fact-based Cuerrier test, the respondent suggests that courts can take judicial notice of the fact that condom use always negates a significant risk of serious bodily harm.

[ 71 ] Before a judge can take judicial notice of a fact, the fact must be shown to be so “notorious” or in modern parlance, “accepted”, that no reasonable person would dispute it: R. v. Find , 2001 SCC 32 , [2001] 1 S.C.R. 863, at para. 48 ; R. v. Spence , 2005 SCC 71 , [2005] 3 S.C.R. 458, at para. 53 . Yet the record here is replete with debate about whether use of a condom alone negates significant risk of serious bodily harm, and the controversy is exacerbated by the rapidly changing state of the science and by the fact-specific nature of risk. Judicial notice is not available here and cannot form the basis for formulating general propositions relating to the factual issue of risk, in the absence of indisputable consensus.

(e) Relationship-based Distinctions

[ 72 ] Another way of narrowing the Cuerrier “significant risk of serious bodily harm” approach to fraud under s. 265(3) ( c ) is to confine it to special relationships. It has been suggested in the commercial context that failure to disclose may amount to fraud when parties have a relationship of trust, quasi-trust or confidence: B. L. Nightingale, The Law of Fraud and Related Offences (loose-leaf). The distinction could conceivably be extended to relationships where one party is vulnerable, like the young often intoxicated women who had sex with Mr. Mabior in this case. This approach assumes that sexual partners who do not ask about STDs bear a risk of exposing themselves to HIV, unless a relationship of trust exists between them.

[ 73 ] The law of fraud in commercial situations provides some support for this approach. Originally, the common law and courts of equity did not regard mere silence, standing on its own without some other form of conduct which induced a false belief in another, as being fraudulent: Twining v. Morrice (1788), 2 Bro. C.C. 326, 29 E.R. 182 ; Conolly v. Parsons (1797), 3 Ves. 625n (Ch.); Walters v. Morgan (1861), 3 De G. F. & J. 718, 45 E.R. 1056 . However, over time, equity recognized specific circumstances warranting a positive duty to disclose material facts, including a relationship of trust, quasi-trust or confidence.

[ 74 ] Against this approach, it can be argued that it narrows “fraud” under s. 265(3) ( c ) too much. Is there a good reason for compelling disclosure to one’s wife but not to a casual date? Where the concern is the contraction of a life-altering disease, the answer is no.

[ 75 ] Historically, fraud capable of vitiating consent to sexual intercourse has not followed the patterns of commercial fraud. The special context of sexual fraud raises unique concerns. This has generally resulted in a narrower approach to fraud, marked by its own peculiar considerations. The relationship between the parties in a particular case has tended not to figure among them. The closest the law has come to a relationship-based approach are the suggestions in Victorian times that wives and prostitutes — for different reasons relating to the mores of the time — could never assert fraud vitiating consent to sexual intercourse: see Clarence . Such ideas strike the modern ear, attuned to equality, as offensive.

(f) The Reasonable Partner Approach

[ 76 ] Yet another possibility is to define “fraud” under s. 265(3) ( c ) as what a reasonable and informed person in the position of the HIV-positive person’s potential partner would expect. The test would be objective, but based on the circumstances of the particular case, including the type of relationship the parties had.

[ 77 ] This approach has the advantage of taking into account the expectations of the parties as they existed on the facts of the case. In this sense, it is fairer than a test that is based on a general norm presumed to apply in all cases. An HIV-positive person may be expected to understand the expectations of his partner in the particular circumstances at hand, and act accordingly.

[ 78 ] This approach also recognizes that allocation of responsibility for protecting against transmission of sexual diseases may vary with relationships and with the passage of time. One of the historic difficulties in formulating a test for fraud vitiating consent to sexual relations has been the variation in social norms and expectations over time. In Victorian times, husbands were held to be incapable of committing fraud vitiating consent, since wives had no right to refuse consent to sexual intercourse with their husbands. In more recent times, some argue that expectations have shifted from the view that an infected person bears all the responsibility for the sexual health of both partners, to the view that each party is responsible for his or her own sexual health. A contextually grounded reasonableness approach would avoid these problems by tailoring the result to the relationship from which the criminal charge arises.

[ 79 ] Again, however, the approach can be criticized. The most telling criticism is that it does not lay down a clear test for fraud, leaving people uncertain as to where the line lies between criminal conduct and non-criminal conduct. Unless the partners have themselves established the lines, the HIV-positive person is left to infer from the nature of the relationship what his potential partner, acting reasonably, would expect.

[ 80 ] This leads to a second criticism — overreach of the criminal law, and its correlative, unfairness to the accused. In the heat and anticipation of the sexual moment, assessments of what a reasonable partner would expect may be mistaken. Is the person then to be criminalized and sent to prison for perhaps years, for what is at its base the result of misjudgment? Such may be the consequence of the reasonable partner expectation test.

(g) An Evolving Common Law Approach

[ 81 ] This leaves the option of building greater certainty into the Cuerrier test by indicating when the significant risk test is met in terms of principle and concrete situations. Such an approach is consistent with the incremental role of common law courts in adapting the contours of an offence to the requirements of justice and practical application.

[ 82 ] The question, building on Cuerrier , is: when do sexual relations with an HIV-positive person pose a “significant risk of serious bodily harm”? In answering this question, the facts as found by the trial judge should be accepted, absent palpable and overriding error. Whether those facts establish a “significant risk of serious bodily harm” is a question of law: see C.A., at para. 37. The terms “serious bodily harm” mean “any hurt or injury, whether physical or psychological, that interferes in a substantial way with the integrity, health or well-being of a victim”: R. v. McCraw , 1991 29 (SCC) , [1991] 3 S.C.R. 72, at p. 88. Of course, actual transmission of HIV constitutes serious bodily harm.

[ 83 ] The courts below answered this question of law differently. The trial judge took the approach that any risk of transmission of HIV, however small, constitutes “significant risk of serious bodily harm” (see reference to the trial judge’s reasons, at C.A. para. 66). The Court of Appeal, by contrast, held that “significant risk” connoted a high risk of transmission of HIV. In its view “significant risk of serious bodily harm” under Cuerrier requires a “significant or high risk” (para. 80) ― the “opposite of evidence of a ‘high probability’ of no infectiousness” (para. 127).

[ 84 ] In my view, a “significant risk of serious bodily harm” connotes a position between the extremes of no risk (the trial judge’s test) and “high risk” (the Court of Appeal’s test). Where there is a realistic possibility of transmission of HIV , a significant risk of serious bodily harm is established, and the deprivation element of the Cuerrier test is met. This approach is supported by the following considerations.

[ 85 ] First, “significant risk of serious bodily harm” cannot mean any risk, however small. That would come down to adopting the absolute disclosure approach, with its numerous shortcomings, and would effectively read the word “significant” out of the Cuerrier test.

[ 86 ] Second, a standard of “high” risk does not give adequate weight to the nature of the harm involved in HIV transmission. “Significant risk” in Cuerrier is informed both by the risk of contraction of HIV and the seriousness of the disease if contracted. These factors vary inversely. The more serious the nature of the harm, the lower the probability of transmission need be to amount to a “significant risk of serious bodily harm”.

[ 87 ] Third, as discussed earlier in considering guides to interpretation, a standard of realistic possibility of transmission of HIV avoids setting the bar for criminal conviction too high or too low. A standard of any risk, however small, would arguably set the threshold for criminal conduct too low. On the other hand, to limit s. 265(3) ( c ) to cases where the risk is “high” might condone irresponsible, reprehensible conduct.

[ 88 ] Fourth, the common law and statutory history of fraud vitiating consent to sexual relations supports viewing “significant risk of serious bodily harm” as requiring a realistic possibility of transmission of HIV. This history suggests that only serious deceptions with serious consequences are capable of vitiating consent to sexual relations. Interpreting “significant risk of serious bodily harm” in Cuerrier as extending to any risk of transmission would be inconsistent with this. A realistic possibility of transmission arguably strikes the right balance for a disease with the life-altering consequences of HIV.

[ 89 ] Fifth, the values of autonomy and equality enshrined in the Charter support an approach to fraud vitiating consent that respects the interest of a person to choose whether to consent to sex with a particular person or not. The law must strike a balance between this interest and the need to confine the criminal law to conduct associated with serious wrongs and serious harms. Drawing the line between criminal and non-criminal misconduct at a realistic possibility of transmission arguably strikes an appropriate balance between the complainant’s interest in autonomy and equality and the need to prevent over-extension of criminal sanctions.

[ 90 ] Finally, interpreting “significant risk of serious bodily harm” as entailing a realistic possibility of transmission of HIV is supported by a number of cases. Apart from the trial reasons in this case, we have been referred to no case holding that “significant” means any risk, however small. To be sure, not many cases have considered the point. But the few that have are worth noting. In R. v. Jones , 2002 NBQB 340, [2002] N.B.J. No. 375 (QL) , the court held that a risk of transmission of hepatitis C between 1.0 and 2.5% was “so low that it cannot be described as significant” (para. 33). And in R. v. J.A.T. , 2010 BCSC 766 () , the trial judge stated that “[a] significant risk means a risk that is of a magnitude great enough to be considered important” (para. 56).

[ 91 ] These considerations lead me to conclude that the Cuerrier requirement of “significant risk of serious bodily harm” should be read as requiring disclosure of HIV status if there is a realistic possibility of transmission of HIV. If there is no realistic possibility of transmission of HIV, failure to disclose that one has HIV will not constitute fraud vitiating consent to sexual relations under s. 265(3) ( c ).

[ 92 ] The test of realistic possibility of transmission proposed in these reasons is specific to HIV. As discussed above, “significant risk” depends both on the degree of the harm and risk of transmission. These two factors vary inversely. A treatable sexually transmitted disease that does not seriously alter a person’s life or life-expectancy might well not rise to the level of constituting “serious bodily harm”, and would also fail to meet the requirement of endangerment of life for aggravated sexual assault under s. 273(1). Where the line should be drawn with respect to diseases other than HIV is not before us. It is enough to note that HIV is indisputably serious and life-endangering. Although it can be controlled by medication, HIV remains an incurable chronic infection that, if untreated, can result in death. As such, the failure to advise a sexual partner of one’s HIV status may lead to a conviction for aggravated sexual assault under s. 273(1) of the Criminal Code . (This said, it may be that with further medical advances, the death rate may decline to the point where the risk of death is virtually eliminated, reducing the offence to sexual assault simpliciter under s. 271(1) of the Criminal Code . Similarly, the day may come when researchers will find a cure for HIV, with the possible effect that HIV will cease to cause “serious bodily harm” and the failure to disclose will no longer fall under the category of fraud vitiating consent for the purposes of sexual assault.)

(4) Realistic Possibility of HIV Transmission

[ 93 ] A review of the case law pertaining to fraud vitiating consent to sexual relations leads to the following general principle of law: the Cuerrier requirement of a “significant risk of serious bodily harm” entails a realistic possibility of transmission of HIV . This applies to all cases where fraud vitiating consent to sexual relations is alleged on the basis of the non-disclosure of HIV-positive status.

[ 94 ] This leaves the question of when there is a realistic possibility of transmission of HIV. The evidence adduced here satisfies me that, as a general matter, a realistic possibility of transmission of HIV is negated if (i) the accused’s viral load at the time of sexual relations was low, and (ii) condom protection was used.

[ 95 ] The conclusion that low viral count coupled with condom use precludes a realistic possibility of transmission of HIV, and hence does not constitute a “significant risk of serious bodily harm” on the Cuerrier test, flows from the evidence in this case. This general proposition does not preclude the common law from adapting to future advances in treatment and to circumstances where risk factors other than those considered in this case are at play.

[ 96 ] In this case, the evidence accepted by the trial judge was that Mr. Mabior had intercourse by vaginal penetration with the four complainants, during which he ejaculated. The accused underwent antiretroviral treatment on a regular basis, but he only wore condoms occasionally. Mr. Mabior’s medical record also shows that he had been successfully treated for gonorrhœa and had no sexually transmitted infections at the time of intercourse with the four complainants other than HIV. Based on this evidence, three factors were relevant to risk: baseline risk for vaginal intercourse when the male partner is HIV-positive; reduction in risk with condom use; and antiretroviral therapy. I will consider each in turn.

[ 97 ] As the Court of Appeal noted, baseline risk of HIV transmission per act of vaginal intercourse with an infected male partner (i.e. the risk of transmission based on the male ejaculating, without a condom, having a normal unreduced viral load) varies from study to study. Dr. Smith’s written report put the risk at 0.05% (1 in 2000) to 0.26% (1 in 384) (p. 4). Ms. McDonald, a public health nurse, testified about the Manitoba Health post-exposure protocol, which puts risk per act at 0.1% (1 in 1000). A systemic review and meta-analysis of 43 publications comprising 25 different study populations, put the risk in high-income countries at 0.08% per sexual act (1 in 1250): M.-C. Boily, et al., “Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies” (2009), 9 Lancet Infect. Dis. 118.

[ 98 ] It is undisputed that HIV does not pass through good quality male or female latex condoms. However, condom use is not fail-safe, due to the possibility of condom failure and human error. Dr. Smith testified that consistent condom protection reduces the risk of HIV transmission by 80%, relying on the widely accepted Cochrane review: S. C. Weller and K. Davis-Beaty, “Condom effectiveness in reducing heterosexual HIV transmission” (2002), 1 Cochrane Database Syst. Rev . CD003255. It was pointed out that the 80% reduction in the transmission risk refers to consistent condom use: the reduction may be larger for consistent and correct condom use, but this has not been verified empirically.

[ 99 ] The Court of Appeal, in applying a “high risk” approach, held that condom use reduces the risk of HIV transmission “below the level of significance” (para. 87). However, in my view, the evidence does not establish that condom protection alone precludes a realistic possibility of transmission, the standard proposed here. According to the expert evidence, the risk might still fall above the “negligible” threshold: Dr. Smith’s report, at p. 6.

[ 100 ] This brings us to the final factor, antiretroviral therapy. As noted by the Court of Appeal, the transmissibility of HIV is proportional to the viral load, i.e. the quantity of HIV copies in the blood. The viral load of an untreated HIV patient ranges from 10,000 copies to a few million copies per millilitre. When a patient undergoes antiretroviral treatment, the viral load shrinks rapidly to less than 1,500 copies per millilitre (low viral load), and can even be brought down to less than 50 copies per millilitre (undetectable viral load) over a longer period of time. This appears to be scientifically accepted at this point, on the evidence in this case.

[ 101 ] Dr. Smith explained in his report that antiretroviral therapy is not a safe-sex strategy, and that “it is highly advisable that persons even with an undetectable viral load who are having sex with more than one partner unfailingly and correctly use a condom” (pp. 5 and 7). The most recent wide-scale study on this issue, relied on by a number of interveners, concludes that the risk of HIV transmission is reduced by 89 to 96% when the HIV-positive partner is treated with antiretrovirals, irrespective of whether the viral load is low or undetectable: M. S. Cohen, et al., “Prevention of HIV-1 Infection with Early Antiretroviral Therapy” (2011), 365 New Eng. J. Med. 493. This evidence indicates that antiretroviral therapy, alone, still exposes a sexual partner to a realistic possibility of transmission. However, on the evidence before us, the ultimate percentage risk of transmission resulting from the combined effect of condom use and low viral load is clearly extremely low ― so low that the risk is reduced to a speculative possibility rather than a realistic possibility.

[ 102 ] In reaching this conclusion, I use low viral load, rather than undetectable viral load, as one of the factors for determining risk. This avoids the evidentiary difficulties associated with establishing an undetectable viral load. Dr. Smith presented evidence that various factors such as common infections and treatment issues may lead to fluctuations in a person’s viral load. So-called “spikes” or “blips” bring the viral load of a person treated with antiretrovirals above the detectable level. Furthermore, detectability depends on the accuracy of ever-developing technology: a viral load that assays do not detect today, might very well be detectable by future assays. Finally, it must be observed that Dr. Smith did not accept, and presented as controversial, the 2008 announcement by the Swiss Federal Commission for HIV/AIDS that an HIV-positive person with an undetectable viral load is not sexually infectious (p. 5). As he noted, this statement is subject to important qualifications, based only on a review of scientific literature, and requires corroborating research.

[ 103 ] This leads to the conclusion that on the evidence before us, the combined effect of condom use and low viral load precludes a realistic possibility of transmission of HIV. In these circumstances, the Cuerrier requirement of significant risk of serious bodily harm is not met. There is no deprivation within the meaning of Cuerrier and failure to disclose HIV status will not constitute fraud vitiating consent under s. 265(3) ( c ) of th e Criminal Code .

(5) Summary

[ 104 ] To summarize, to obtain a conviction under ss. 265(3)( c ) and 273, the Crown must show that the complainant’s consent to sexual intercourse was vitiated by the accused’s fraud as to his HIV status. Failure to disclose ( the dishonest act ) amounts to fraud where the complainant would not have consented had he or she known the accused was HIV-positive, and where sexual contact poses a significant risk of or causes actual serious bodily harm ( deprivation ). A significant risk of serious bodily harm is established by a realistic possibility of transmission of HIV. On the evidence before us, a realistic possibility of transmission is negated by evidence that the accused’s viral load was low at the time of intercourse and that condom protection was used. However, the general proposition that a low viral load combined with condom use negates a realistic possibility of transmission of HIV does not preclude the common law from adapting to future advances in treatment and to circumstances where risk factors other than those considered in the present case are at play.

[ 105 ] The usual rules of evidence and proof apply. The Crown bears the burden of establishing the elements of the offence — a dishonest act and deprivation — beyond a reasonable doubt. Where the Crown has made a prima facie case of deception and deprivation as described in these reasons, a tactical burden may fall on the accused to raise a reasonable doubt, by calling evidence that he had a low viral load at the time and that condom protection was used.

B. Application

[ 106 ] With respect to the four counts before us, the complainants all consented to sexual intercourse with the accused. Each of the complainants testified that they would not have had sex with the accused had they known that he was HIV-positive. The only issue is whether their consent was vitiated because he did not tell them that he had HIV.

[ 107 ] The trial judge found the accused guilty of aggravated sexual assault on the four counts where it was established that his viral load was not undetectable or no condom was used. The Court of Appeal set aside the convictions on the basis that either an undetectable viral load or condom protection would suffice.

[ 108 ] As set out above, at this point in the development of the common law, a clear test can be laid down. The absence of a realistic possibility of HIV transmission precludes a finding of fraud vitiating consent under s. 265(3) ( c ) of the Criminal Code . In the case at hand, no realistic possibility of transmission was established when the accused had a low viral load and wore a condom. It follows that the appeal should be allowed insofar as the decision of the Court of Appeal conflicts with this conclusion.

[ 109 ] The accused had a low viral load at the time of intercourse with each of S.H., D.C.S. and D.H., but did not use a condom. Consequently, the trial judge’s convictions on these counts should be maintained. This leaves K.G. The trial judge convicted on the ground that, although the accused used a condom at the time of the encounter, his viral load “was not suppressed” (para. 128). As discussed, the combination of a low viral load — as opposed to an undetectable viral load — and of condom use negates a realistic possibility of transmission, on the evidence in this case. The record shows that the accused’s viral load was low at the time of sexual relations with K.G. When combined with condom protection, this low viral load did not expose K.G. to a significant risk of serious bodily harm. The trial judge’s conviction on this count must be reversed.

[ 110 ] I would allow the appeal in part and restore the convictions in respect of the complaints by S.H., D.C.S. and D.H. I would dismiss the appeal in respect of the complaint by K.G.

Appeal allowed in part.

Solicitor for the appellant: Attorney General of Manitoba, Winnipeg.

Solicitor for the respondent: Legal Aid Manitoba, Winnipeg.

Solicitors for the interveners the Canadian HIV/AIDS Legal Network, the HIV & AIDS Legal Clinic Ontario, Coalition des organismes communautaires québécois de lutte contre le sida, the Positive Living Society of British Columbia, the Canadian AIDS Society, the Toronto People With AIDS Foundation, the Black Coalition for AIDS Prevention and the Canadian Aboriginal AIDS Network: Cooper & Sandler, Toronto.

Solicitors for the intervener the British Columbia Civil Liberties Association: McCarthy Tétrault, Vancouver.

Solicitors for the intervener the Criminal Lawyers’ Association of Ontario: Schreck Presser, Toronto.

Solicitors for the intervener Association des avocats de la défense de Montréal: Poupart, Dadour, Touma et Associés, Montréal.

Solicitors for the intervener Institut national de santé publique du Québec: Desrosiers, Joncas, Massicotte, Montréal.