Claim and Category - Compulsive Sexual Behaviour, the Courts, and the Framework Canada Never Built

“Sex addiction” is not a medical diagnosis. A related condition - compulsive sexual behaviour disorder - was formally recognized in 2022, but as an impulse control disorder, which does not mean a person cannot control their actions. In Canadian courts, a claim of sexual addiction, or enrolment in a treatment program, is often treated as a reason for leniency - frequently without anyone testing whether the claim holds up. This article explains how a courtroom argument came to outrun the science it borrows from, and how Canada could close the gap using tools its highest court already built.

In 2017, the expert objection to the “sex addiction” defence was that no recognized diagnosis existed. In 2022, a diagnosis arrived - compulsive sexual behaviour disorder, classified as an impulse control disorder - a category that does not support a narrative of powerlessness. The courtroom claim carried on unchanged.

Forensic psychiatry’s own survey of the case law found courts treating the claim as a mitigating factor in ten of fifteen cases while almost never testing its legitimacy, and concluded that the risks of using the condition in legal proceedings outweigh the benefits.

Canada’s only survey of its own courts dates from 2017 - the year before the diagnosis existed - and found judges reading a defendant’s support-group attendance as rehabilitative potential while sidestepping whether the condition is real. It has never been updated.

This article argues for a testing framework, not against treatment. Nothing here suggests that people experiencing compulsive sexual behaviour are dangerous, or that seeking help is anything but sound.

Introduction

In December 2017, with the MeToo reckoning in full force and a procession of public figures announcing themselves into residential treatment, the Associated Press canvassed the experts on “sex addiction.” The verdict, carried in Canada by CTV News, was close to unanimous: whatever was wrong, it was not an officially recognized psychiatric diagnosis - even as the same clinicians acknowledged that compulsive sexual behaviour can genuinely upend lives.

Nine years on, that article reads as a time capsule with an unusual property. Its central objection has since been answered. Its central worry has not.

In January 2022, a diagnosis came into force. The World Health Organization’s ICD-11 recognizes compulsive sexual behaviour disorder: a persistent pattern of failure to control intense, repetitive sexual impulses or urges, sustained over six months or more. It sits among the impulse control disorders, alongside kleptomania, pyromania, and intermittent explosive disorder - a deliberate placement, made because the evidence was insufficient to classify it with the addictive disorders. The courtroom claim absorbed none of this. It runs today as it ran in 2017: an addiction, a disease, a treatment program, a changed person.

That is the finding of this article. The diagnostic ground shifted - a recognized condition arrived where there had been none - and the defence built on the old ground did not move with it. In 2017 the objection was that no diagnosis existed. By 2022 one did, but not one shaped to excuse.

A defence that runs the same either way was never resting on the diagnosis. It is rhetoric wearing a clinical vocabulary - and a court with no framework for testing it cannot tell the difference.

Three things this article does not argue. It does not argue that people who experience compulsive sexual behaviour are dangerous; most are a risk to no one. It does not argue that treatment is worthless or that seeking help is suspect; for many people it is the right decision. And it names no litigants: everything here draws on published case law and peer-reviewed analysis, read for pattern.

The Diagnosis Arrived

Start with what the 2017 experts could not have cited, because it did not yet exist.

Compulsive sexual behaviour disorder entered the eleventh revision of the International Classification of Diseases, in force since January 2022. The classification decision matters as much as the recognition. The World Health Organization placed the condition among impulse control disorders rather than addictive disorders because the evidence for an addiction model was insufficient - a placement the field’s own literature describes as deliberate (both terms are defined in the Glossary). An impulse control disorder names a repeated failure to resist urges. It does not deliver the disease-of-powerlessness narrative that the word “addiction” imports. The label medicine adopted and the label counsel reach for are not the same label, and the distance between them is the whole argument.

The American manual went further, by staying silent. The DSM-5 declined a formally proposed diagnosis of hypersexual disorder in 2013, and the DSM-5-TR contains nothing equivalent today. The forensic literature records something pointed about that refusal: alongside insufficient evidence, a primary reason the American Psychiatric Association’s Board of Trustees declined the diagnosis was concern about its potential misuse in legal settings. The manual’s silence is not an oversight. It is, in part, a judgment about the courtroom.

There is more. The diagnostic guidelines that accompany the new category do not merely permit scrutiny of the self-applied label - they require it. The authors of the ICD-11 proposal note that careful attention must be paid to people who self-identify as “sex addicts” or “porn addicts,” because on examination such individuals may not actually exhibit the clinical features of the disorder. In other words, the classification’s own framers instruct that the self-label be tested against criteria rather than taken at face value - which is precisely the step the courtroom claim skips.

The field’s own professional bodies had reached their position even earlier. In November 2016 - a full year before the Associated Press piece - the American Association of Sexuality Educators, Counselors and Therapists stated that it “does not find sufficient empirical evidence” to support classifying sex addiction or porn addiction as a mental health disorder, and does not find sex-addiction training and treatment methods adequately informed by accurate human sexuality knowledge. That second finding is worth holding onto; this article returns to the training industry below.

By 2022, then, the 2017 objection was obsolete - a recognized diagnosis existed. But the diagnosis that arrived is the wrong shape for the defence advanced in its name.

The Courtroom Narrative

What the case law shows

The most systematic account of how courts actually handle the claim comes from forensic psychiatry’s own legal journal. In 2022, Nassif and colleagues searched the LexisNexis database for cases in which sexual addiction figured in legal proceedings. The term surfaced 369 cases; fifteen met their criteria of expert involvement and a diagnosis or substantive discussion. Ten were criminal - nine of them proceedings against accused sexual offenders - four were civil, and one arose in family law.

In ten of the fifteen, the court perceived sexual addiction as a mitigating factor.

More telling than the count is what the courts did not do. Neither this review nor the earlier American survey it builds on found courts subjecting the testimony to evidentiary scrutiny to determine whether the diagnosis was legitimate. Admissibility varies by jurisdiction, and in practice the question was mostly never reached. What moved courts instead were factors orbiting the claim: the number and nature of offences, comorbid conditions, perceived risk - and “degree of remorse as reflected by investment in treatment.” Hold that phrase. The research in the next section returns to it.

The review’s conclusion is blunt for a professional journal. Whichever diagnostic category the condition might land in has little bearing on risk or on what treatment would achieve. Until those questions are resolved, the authors write, the risks of using the condition in trial proceedings outweigh the proposed benefits - and forensic experts should make the debates, concerns, and gaps in knowledge clear to the courts regardless of which side retains them. They add that the condition would almost certainly not sustain an insanity defence, since it does not bear on a person’s capacity to understand the nature or wrongfulness of their conduct.

The label cuts both ways

The American picture beneath that conclusion is not one of uniform leniency. The first review of these cases, published by Ley and colleagues in 2015, found the “sex addiction” label cutting both ways: raised by defendants to mitigate, but also wielded by plaintiffs and prosecutors as an aggravating marker of dangerousness or risk. The label is not reliably a shield. It is a rhetorical instrument that whichever party finds it useful can pick up - which is precisely what one would expect of a claim doing persuasive rather than diagnostic work.

Canada, surveyed once

Canada has been surveyed exactly once. Stephanie Montgomery-Graham, then at Western University, sampled Canadian court, tribunal, and arbitral decisions through the Quicklaw database and published her findings in 2017. The pattern she documented was subtler than acceptance or rejection: courts largely declined to engage the validity question at all. Where the claim gained traction in criminal matters, the most common outcome was sentencing mitigation - and judges frequently read a defendant’s attendance at support groups such as Sex Addicts Anonymous as evidence of rehabilitative potential, an acknowledgment of the problem and a step toward recovery. That survey was published the year before the diagnosis existed, and it has never been updated. Canada’s only map of its own courtroom practice predates the territory.

A single reported family-law decision shows the mechanism at close range. In a 2012 New Brunswick case - reported, in keeping with the practice for family matters, only by the parties’ initials - a marriage broke down after one spouse discovered the other engaged in online sexual activity. During the litigation, a psychologist gave the opinion that the spouse’s sexually addictive behaviour was poor role modelling for the children. The court ordered that the parent not leave the children unsupervised or exposed, directly or indirectly, to sexual materials or pornography - yet still found both parents entitled to a shared parenting arrangement. The case is instructive precisely because nothing in it went wrong by the standards of the day: a clinical characterization of “sex addiction” entered through an expert, shaped a protective term, and coexisted with shared parenting, all without the court being asked to test whether the characterization rested on a recognized diagnosis. It is the aggregate pattern Montgomery-Graham found, visible in a single file.

What the Research Says About the Claim Itself

Three findings, all postdating or unavailable to the 2017 debate, bear directly on what a court can responsibly do with the claim.

1. The mechanism is not established

The forensic review notes that studies using neuropsychological and psychometric testing found no significant differences in executive functioning or impulsivity between hypersexual men and controls. A defence that says he could not stop himself presupposes a deficit that measurement has not located.

2. Self-diagnosis is unreliable

A research programme led by Grubbs and colleagues, synthesized in a 2019 systematic review and meta-analysis, found that feelings of being addicted to pornography are in many cases better explained by moral incongruence than by the behaviour itself. Moral incongruence is the collision between a person’s behaviour and their own moral beliefs about it (see Glossary). This is not only a research finding; the ICD-11 diagnostic guidelines themselves exclude it, stating that the disorder should not be diagnosed from psychological distress related to moral judgments or disapproval, and that distress arising from sexual behaviour alone does not warrant the diagnosis.

A person who declares themselves an addict may be describing shame, not a disorder. And someone who meets clinical criteria may sincerely reject the label. Either way, the diagnosis-as-self-report that features in proceedings - he has admitted he is an addict; he has taken responsibility - carries far less information than it appears to.

3. Treatment behaviour signals little on its own

The largest international study of the disorder, spanning 82,243 participants in forty-two countries, found that only fourteen per cent of those identified as experiencing it had ever sought treatment. (Prevalence estimates for the disorder vary with the instrument and threshold used - the ICD-11’s more restrictive criteria yield lower figures than broad screening tools - so the point that matters here is the treatment-seeking proportion, not the headline rate.) Presence in a program is therefore no marker of severity, and absence is no marker of health. None of this is an argument against seeking help: many people report real benefit from structured treatment and peer support, and that benefit is reason enough to pursue it. The point is narrower and strictly evidentiary - what helps a person privately does not, without more, prove anything a court needs to find. Now recall what courts read investment in treatment as: remorse and rehabilitative potential. A signal the population data marks as nearly uninformative is functioning, in the case law, as a proxy for the thing courts most want to know.

And where the program is a peer fellowship, even the fact of participation is beyond reach. Twelve-step fellowships for sexual behaviour operate on structural anonymity: membership is unrecorded, attendance is unverifiable, and what is said in a meeting is reviewable by no one outside it. Those protections are deliberate, and they serve people who genuinely need them. They also mean that, evidentially, a claim of participation is an assertion and nothing more. It cannot be corroborated, and it cannot be tested. Montgomery-Graham’s finding is the live demonstration: Canadian courts were reading exactly this unverifiable attendance as rehabilitative potential.

The three findings can be set directly against the assertions they answer - the courtroom claim on the left, what the evidence actually supports on the right:

Mitigation, risk assessment, proof of rehabilitation: each rests on an input that the research now marks as unestablished, unreliable, or untestable.

There is a serious argument on the other side, and it deserves stating at its strongest. Courts, the argument runs, are not laboratories; they decide under uncertainty every day, and a person who names a problem and seeks help has done something meaningfully different from one who does neither. Waiting for definitive science before crediting any effort at change would punish exactly the conduct the system wants to encourage, and would leave judges paralysed on questions they are obliged to resolve. That is a fair account of the adversarial task, and it is why this article does not argue that treatment participation should count for nothing. The response is narrower: the objection is not that courts credit effort, but that they credit an unverified claim of effort as though it were established fact. A person who genuinely engages treatment can almost always show it - through clinical records, a treating professional, a verifiable course of care. What the evidence cannot support is weight assigned to the bare assertion, or to attendance at a venue where attendance is by design unprovable. Crediting effort is sound. Crediting the label without testing it is the failure. The gap between those two is the whole of what this article asks courts to close.

The Canadian Framework That Doesn’t Reach It

Canadian law does not lack tools for exactly this problem. It has simply never issued guidance on applying them to it.

The Supreme Court of Canada built the machinery decades ago. R. v. Mohan sets the threshold for expert evidence: relevance, necessity, the absence of an exclusionary rule, a properly qualified expert. R. v. J.-L.J. directs special scrutiny at novel science. White Burgess imposes on every expert a duty of independence and impartiality that is a precondition of admissibility. A contested diagnosis, advanced by a retained expert, resting on disputed science, is close to the fact pattern this jurisprudence was made for. No published Canadian guidance applies it to compulsive sexual behaviour claims.

And the forum where the claim now does much of its work is the one where the machinery engages least. Criminal proceedings at least stage the claim through an expert, whose testimony can in principle be challenged; family proceedings frequently receive it with no expert at all. In family proceedings generally, such claims surface not through expert testimony but in the ordinary documents of the file. An affidavit deposes to program attendance; a counsellor’s letter is offered as proof of insight; a treatment certificate is exhibited to show change; a plan of care recites a recovery narrative; submissions recast a pattern of concealment as a managed condition now under treatment. Each enters on a balance of probabilities, and where no opinion evidence is formally tendered, the Mohan framework never activates - so material a criminal court would scrutinize passes into a best-interests determination untested.

The gap is not that the principles are unavailable in family court; it is that nothing prompts their use. A counsellor’s letter asserting “insight” or “recovery” is, in substance, an opinion about a clinical state - and a judge retains the discretion to treat it as such: to ask what qualifies its author to offer that opinion, whether it rests on a recognized diagnosis or on the writer’s acceptance of the party’s self-report, and what independent basis supports it. The same questions Mohan asks of a retained expert can be asked of an affidavit exhibit, without any new rule - a self-represented respondent, or the court itself, need only raise them. What is missing is not the authority but the cue to apply it, which is where bench guidance (below) would do its work. Montgomery-Graham’s finding that Canadian judges sidestep the validity question suggests the gap does not close even when the subject does surface.

Parliament recently modernized half of this picture. The 2021 amendments to the Divorce Act define family violence broadly - physical, sexual, and psychological abuse, financial abuse, coercive and controlling patterns - and include a child’s direct or indirect exposure to such conduct. Courts must weigh it in every best-interests determination. What Parliament built is a lens for conduct alleged against a parent. What it did not build is any tool for testing the narrative a parent offers in response: the diagnosis, the program, the recovery. The statute can now see coercion. It still cannot test rehabilitation.

Beneath the claims sits an industry layer that courts rarely examine. Certifications in “sex addiction therapy” issue from private training organizations, not from health regulators, and AASECT’s 2016 position went precisely to the adequacy of that training. Where provinces regulate psychotherapy, they regulate practitioners - not the validity of the model a practitioner advertises. A court that weighs a treatment credential without asking who issued it imports the industry’s self-assessment as evidence.

Why the Untested Claim Matters

It would be possible to read all of this as a technical quarrel about evidence. It is not. In a family proceeding, the claim is not weighed in the abstract - it is weighed to decide what a child will be exposed to, and on what terms. That is where an untested assertion stops being an evidentiary curiosity and becomes a decision about a child’s daily life.

The stakes do not depend on any claim that the diagnosis causes harm; the evidence does not support that, and this article does not make it. The stakes lie one level down, in conduct and exposure. Where a pattern of concealed conduct is real, a child in the household may be exposed to it directly or through its traces - and exposure of that kind is a documented harm vector in the child-development and child-protection literature, independent of any label attached to the adult. The other adult is affected too: a distinct clinical literature documents that the discovery of concealed sexual betrayal can itself produce trauma responses in a partner - intrusive thoughts, hypervigilance, acute distress - and a household absorbing that shock is a household under strain around any children in it. A companion analysis on this site examines that partner and household exposure in full. The point here is narrower: when a court accepts a recovery narrative untested, it is not merely crediting an unproven claim. It is settling, on that unproven basis, what a child is exposed to.

And Canadian law already supplies the tool. The best-interests test does not wait on new diagnostic categories or on research yet to mature. Section 16 of the Divorce Act makes the best interests of the child the only consideration in a parenting order, and directs the court to weigh, among the child’s circumstances, any family violence and its impact - including a child’s exposure to it - and each parent’s willingness to support the child’s development. A child’s exposure to a parent’s conduct is therefore already, expressly, something the statute tells a court to weigh, on the evidence actually before it. What the untested claim does is short-circuit that weighing - substituting an assertion of recovery for the assessment the child is owed. The recommendations that follow are ways to restore it.

A Framework Worth Building

Each recommendation addresses a documented gap identified above.

1. Commission the Canadian update. The only survey of how Canadian courts treat these claims dates from 2017, before the diagnosis existed. A law foundation, the federal Department of Justice, or academic funders should commission its successor, covering the post-2022 case law. Addresses: a map of Canadian practice that predates the territory.

2. Issue bench guidance distinguishing diagnosis from defence. Judicial education in Canada runs through the National Judicial Institute, the judge-led body responsible for the overall coordination of judicial education and a primary provider of it, with professional-development requirements for superior court judges set by the Canadian Judicial Council. An electronic bench book or education module from the NJI should set out what exists (an ICD-11 impulse control disorder), what does not (any DSM equivalent, declined partly out of concern for courtroom misuse), and what the professional bodies have said. A judge cannot weigh a contested category she has never been briefed on. Addresses: courts sidestepping the validity question.

3. Apply the existing gatekeeping to this claim. Where opinion evidence is tendered, Mohan, J.-L.J., and White Burgess already supply the test. Experts on any side should be expected to disclose the classification dispute, the unresolved impulse-control findings, and the outcome-evidence gap - the forensic literature’s own instruction to its members, made a threshold expectation of Canadian courts. Addresses: testimony entering without evidentiary scrutiny.

4. Treat participation claims as assertions absent corroboration. Where anonymity is structural, attendance cannot be verified and content cannot be reviewed; a claim of participation is evidence of a stated intention, nothing further. Where treatment is clinical, it can be evidenced through records or the treating clinician, with ordinary safeguards. Weight should follow verifiability - not, as the 2017 Canadian survey found, flow automatically from the bare fact of attendance. Addresses: treatment investment functioning as a proxy for remorse.

5. Ask who issued the credential. Courts weighing a treatment program or an expert’s specialization should establish whether the credential behind it comes from a health regulator or a private training body. Regulators and consumer-protection law should reach advertising that presents private certification as clinical qualification. Addresses: the industry layer beneath treatment claims.

6. Use the category that exists. This is a point about terminology, not about changing any legal test or outcome. Reasons for judgment shape later judgments, and the words they use travel. Where the evidence supports a finding at all, the finding is compulsive sexual behaviour disorder, an impulse control disorder - not “addiction,” a word that imports the powerlessness the science declined to adopt. Nothing in the substantive law changes; only the label is made accurate. Addresses: the vernacular doing doctrinal work.

None of this requires a court to disbelieve anyone. It requires only that a claim with legal consequences be treated like every other claim with legal consequences: as something to be tested.

The Stakes, and the Path

Between 2017 and today, almost everything about this condition changed. A diagnosis was recognized, and recognized in a form that undercuts the powerlessness the courtroom claim depends on. The research on self-diagnosis, on treatment-seeking, and on the mechanism the defence assumes all moved. The one thing that did not move is the claim itself, still doing persuasive work in rooms that decide sentences and, more consequentially, decide what a child lives with.

The gap this leaves is not a gap in the law. Canada’s highest court built the gatekeeping decades ago, and Parliament wrote a best-interests test broad enough to weigh a child’s exposure to a parent’s conduct. The tools exist. What is missing is the prompt to use them - the bench guidance, the updated survey, the habit of asking of a recovery narrative the same questions the law asks of every other claim. Those are buildable, and none of them requires resolving the science first or accepting any contested claim about who is dangerous. They require only that the system stop treating an untested assertion as a finding. A claim that has outrun its evidence for nine years can be asked, at last, to prove itself. And the reason to ask is not abstract: at the end of an untested claim is a child, living every day with whatever the court declined to examine.

Developments to Watch

The independent long-term evaluation of the German prevention model sites runs to 2026; its results will bear on what any court can be told about treatment outcomes.

Compulsive sexual behaviour disorder remains under active study, and any movement in ICD-11 revision or a future DSM text revision would change the classification ground this article describes.

Canadian decisions have begun to reference the ICD-11 terminology since 2022, though no systematic review of the post-2022 case law yet exists. Building that record is the raw material for the updated survey recommended above; until it is done, the extent to which Canadian courts have absorbed the reclassification is itself unknown.

A companion analysis of the surveillance gap - what Canada measures about child maltreatment and what it has never counted - is published separately on this site.

Glossary

Compulsive sexual behaviour disorder (CSBD) - The diagnosis recognized by the World Health Organization’s ICD-11, in force since January 2022. It describes a persistent pattern of failure to control intense, repetitive sexual impulses or urges, sustained over six months or more. It is the clinical term; “sex addiction” is not.

“Sex addiction” - A popular, contested label, not a recognized diagnosis in either the ICD-11 or the American DSM-5-TR. It imports an assumption of powerlessness drawn from the addiction model - the assumption the formal classification specifically declined to adopt. The gap between this term and CSBD is the subject of this article.

Impulse control disorder - The category in which the ICD-11 places CSBD, alongside kleptomania, pyromania, and intermittent explosive disorder. It names a repeated failure to resist urges. It does not, in itself, establish that a person could not have acted otherwise.

Addictive disorder - A separate ICD-11 category, covering substance and behavioural addictions. CSBD was deliberately not placed here, because the evidence for an addiction model was judged insufficient. The distinction matters legally: “addiction” carries an implication of diminished control that “impulse control disorder” does not.

Mitigating factor - In sentencing or a best-interests determination, a consideration that weighs in a person’s favour. In the case law surveyed here, courts frequently treated a claim of sexual addiction, and investment in treatment, as mitigating - often without testing whether the underlying diagnosis was valid.

Moral incongruence - The distress that arises when a person’s behaviour conflicts with their own moral beliefs about that behaviour. Research finds that self-reported “addiction” to pornography is, in many cases, better explained by this conflict than by the behaviour itself - which makes self-diagnosis an unreliable signal in both directions.

Read Further

The full research and structural analysis are available in: