The criminal justice system can be daunting, but you don’t need to go through it alone. Our Criminal lawyers are here to guide you every step of the way.
Contact Our Firm
Joseph A. Neuberger and Michael P. Bury, Neuberger & Partners LLP, Criminal Lawyers Toronto
For nearly a decade, the Criminal Code prohibited the non-consensual distribution of intimate images while leaving a significant gap at the centre of the offence. Effective July 18, 2026, that gap has been addressed. The amended definition of “intimate image” in section 162.1(2) now expressly encompasses realistic sexual deepfakes. This article examines the nature of the amendment, the jurisprudence that precipitated it, and the boundaries of its application.
01 • THE PRIOR LAW AND ITS LIMITATIONS
Section 162.1 was introduced into the Criminal Code in 2015 through the Protecting Canadians from Online Crime Act. The provision arose in response to the deaths of Rehtaeh Parsons and Amanda Todd, two adolescents who died by suicide following the circulation of intimate images and sustained online harassment. Public concern generated by these events prompted Parliament to establish the non-consensual distribution of intimate images as a discrete offence, rather than leaving prosecutors to rely on offences of general application.
The offence itself was drafted broadly. It captured the publication, distribution, transmission, sale, making available, or advertisement of an intimate image of a person, where the accused knew that the person depicted had not consented or was reckless as to the absence of consent. The difficulty lay not in the conduct elements of the offence but in its definitional foundation.
Under the prior definition, an “intimate image” meant a visual recording of a person — a photographic, film, or video recording — depicting the person nude or engaged in explicit sexual activity, made in circumstances giving rise to a reasonable expectation of privacy. The phrase “visual recording” presupposed that a genuine image of the individual had been captured. A synthetic image produced by software had not been recorded of any actual person. Consequently, when courts were required to determine whether an artificially generated depiction of a real individual satisfied the definition, they encountered a recurring interpretive obstacle.
Two decisions rendered in 2025 and 2026 rendered the limitation untenable. In R. v. Kapoor (Ontario, 2025), the court characterised the sharing of an AI-generated nude depiction of a spouse as morally reprehensible but concluded that the conduct fell outside section 162.1, on the basis that the image had been generated rather than recorded. In R v MSK, 2026 NSPC 12, the accused obtained ordinary clothed photographs of several complainants from social media, employed artificial intelligence software to fabricate realistic nude depictions, and subsequently distributed those depictions. Distribution and the absence of consent were not in dispute; the sole question was whether the fabricated images constituted “intimate images” within the meaning of the section. The court held that they did not amount to “visual recordings of a person,” declined to extend the statute through judicial interpretation, and expressly invited Parliament to revise the definition rather than accommodate deepfakes within a framework designed for a different purpose.
It was against this backdrop that the amendment was enacted.
02 • THE LEGISLATIVE AMENDMENT
The amendment was enacted through Bill C-16, the Protecting Victims Act (S.C. 2026, c. 19). The Act received Royal Assent on June 18, 2026, and the majority of its Criminal Code provisions, including the amendments concerning intimate images, came into force on July 18, 2026 — the thirtieth day following Royal Assent.
The central feature of the amendment is a revised section 162.1(2), which now comprises two branches.
An intimate image continues to include a visual recording of a person, made by any means and including a photographic, film, or video recording, in which the person is nude, is exposing their sexual organs, or is engaged in explicit sexual activity, made in circumstances giving rise to a reasonable expectation of privacy that the person retains at the time the offence is committed. This branch has been expanded to capture depictions in which the person is “nearly nude,” in addition to those depicting full nudity.
The definition is extended to a visual representation made by any electronic or mechanical means — with the statute expressly identifying artificial intelligence software — that depicts an identifiable person as nude, nearly nude, exposing their sexual organs, or engaged in explicit sexual activity, where the depiction is likely to be mistaken for a genuine recording of that person.
The limiting conditions attached to the second branch are of considerable importance. The provision applies to an identifiable person, and not to a generic or fictional figure. It applies, moreover, only where the fabrication is sufficiently realistic that it is likely to be mistaken for a genuine recording of the individual depicted. An obviously artificial or stylized rendering is therefore to be distinguished from a photorealistic fabrication intended to be received as authentic; the provision is directed at the latter. The requirement of a reasonable expectation of privacy that governs the recording branch does not apply in identical terms to the second branch, a distinction that reflects the fact that a fabricated image is invented rather than captured in a private context.
03 • RELATED PROVISIONS OF THE REFORM
The revised definition is accompanied by several related amendments effected by Bill C-16.
Section 162.1(1.1) now establishes a separate offence of threatening to publish or distribute an intimate image, including a deepfake, with the intent to intimidate or to be taken seriously, irrespective of whether the image is ultimately distributed. Previously, a threat of this nature required prosecution under extortion or other provisions of general application. The amendment addresses a practical limitation that had frequently arisen in sextortion matters.
The maximum term of imprisonment for the distribution offence on indictment has been increased from five years to ten years. A higher maximum of fourteen years continues to apply where the accused knew, or ought to have known, that an aggravated sexual assault was being, or had just been, committed against the person depicted at the time the underlying image was made. Both the distribution offence and the new threat offence remain hybrid offences, permitting the Crown to proceed by indictment or by summary conviction according to the gravity of the matter.
04 • MATTERS UNAFFECTED BY THE AMENDMENT
Precision on this point is warranted, as certain commentary following the reform suggested that deepfake abuse became unlawful only in 2026. A substantial proportion of such conduct was already criminal.
Since 2015, the distribution of a genuine intimate image without consent has constituted an offence. The surreptitious observation or recording of a person in a private setting has long been addressed by the voyeurism offence under section 162. The demand for money, images, or other advantage under threat of exposure has consistently been prosecutable as extortion, which carries a maximum penalty of life imprisonment. Where the person depicted is under the age of 18, the child pornography provisions under section 163.1 may apply, including to artificially generated material and including where the person responsible is themselves a minor, subject to the application of the Youth Criminal Justice Act to young offenders.
The more accurate characterisation of the 2026 reform is therefore narrower than the proposition that deepfakes have become unlawful. What the amendment achieves is that the specific offence of non-consensual distribution of intimate images now reaches realistic sexual deepfakes on its own terms, and that the threat to distribute such an image now constitutes a discrete offence.
The public-good defence under section 162.1(3) remains unchanged. No person may be convicted under the section where the conduct serves the public good and does not extend beyond what serves the public good — a defence of narrow scope that, in practice, seldom operates as broadly as might be assumed.
05 • SIGNIFICANCE OF THE AMENDMENT
The practical significance of the amendment is most readily apparent through the decisions that revealed the deficiency. Under the prior definition, an accused in circumstances analogous to those in MSK could distribute fabricated depictions of real individuals, occasion substantial and documented harm, and nonetheless fall outside the very offence that appears designed to address such conduct. The documented consequences of this form of abuse include loss of employment, expulsion from educational institutions, post-traumatic stress, and, in the most severe instances, suicide. The harm was never speculative; the statutory language had simply failed to keep pace with technological developments.
Generative image tools entered widespread use in approximately 2023, and the incidence of deepfake abuse in educational and workplace settings increased markedly thereafter. A number of provinces responded through civil measures, and certain jurisdictions, including Manitoba, amended their provincial statutes to capture fabricated intimate images in advance of the federal reform. The federal amendment now furnishes prosecutors with a criminal instrument aligned with the manner in which this abuse ordinarily occurs — not through a concealed recording device, but through an image obtained from a publicly accessible profile and processed by an artificial intelligence model.
06 • CONCLUSION
Reduced to its essentials, the reform consists of a single precisely directed amendment supported by several ancillary provisions. Section 162.1(2) previously protected individuals only against the misuse of images genuinely recorded of them. It now also protects identifiable persons against realistic, fabricated depictions likely to be mistaken for authentic recordings, and it renders the threat to distribute such a depiction an offence in its own right, whether or not distribution ultimately occurs.
For complainants, the amendment provides a clearer avenue toward a criminal complaint that corresponds to the conduct in question. For those inclined to advance the position that an artificially generated image does not depict the actual individual, that argument is considerably less tenable than it was prior to the reform. More generally, the amendment illustrates the characteristic way criminal law responds to emerging technology: not by anticipating it, but by adapting to it, frequently only after harm has occurred.
This article reflects the law as amended and in force on July 18, 2026, and is provided as general information only. It does not constitute legal advice, and no solicitor-client relationship is created by its publication. Any person involved in a specific matter should obtain advice from counsel concerning their own circumstances.
