A parliamentary committee has strengthened Canada’s proposed law against sexual deepfakes, expanding its reach to cover “nearly nude” images after experts flagged a significant gap in the original draft.
The House of Commons justice committee approved amendments to Bill C-16 last week, responding to concerns that the legislation as written would have missed a growing category of AI-generated abuse particularly images produced by Elon Musk’s Grok chatbot that spread widely on the X platform earlier this year. Those images, which typically depicted women in digitally altered see-through bikinis, fell into a grey area the original bill simply did not address.
As initially drafted, Bill C-16 would have made it a criminal offence to non-consensually share images depicting nudity, exposed genitalia, or explicit sexual activity. But legal experts who testified before the committee argued that standard was already being outpaced by technology. Images that stop just short of full nudity yet are created deliberately to sexualize, humiliate, or harm would have slipped through unscathed.
Conservative MP Andrew Lawton brought the issue to a head, tabling an amendment to expand the definition to include images where a person is nude or “nearly nude.” He said the change was informed by both expert testimony and the personal experience of someone close to him.
“We are seeing with the advancement of technology these very sophisticated and in some cases quite traumatizing assaults taking place,” Lawton told the committee. “It simply just ensures that a small technicality is not excluding something that I think this law intends to capture.”
Liberal MP Patricia Lattanzio, parliamentary secretary to Justice Minister Sean Fraser, backed the amendment, saying it “clarifies the scope of the offence, aligns with evolving case law, and responds to the needs of the victims.”
Not everyone was on board. Bloc Québécois MP Rhéal Fortin raised concerns that the phrase “nearly nude” lacked a precise legal definition an ambiguity, he argued, that could create problems down the line in prosecution.
Beyond expanding what kinds of images are covered, another approved amendment inserts an explicit reference to artificial intelligence software in the bill’s definition of “intimate image.” Lawton framed it as future-proofing the legislation against the rapid pace of technological change.
“We’re trying to ensure that we don’t end up having to come back to the drawing board because this fails to capture the technologies that we’re dealing with here,” he said.
One amendment that did not make the cut came from NDP MP Leah Gazan, who had pushed for even broader language targeting images that depict women and children in sexualized or degrading situations such as transparent swimwear or imagery depicting physical violence like bruising and blood. Gazan argued the images constitute acts of violence designed purely to humiliate.
“We cannot leave open loopholes in this legislation that perpetrators can exploit to evade justice,” she said.
The committee also approved two additional changes put forward by Lawton. The first would increase maximum penalties in cases where the deepfake material depicts sexual assault. The second introduces a 48-hour window for platforms to remove flagged content a provision designed to put real accountability on tech companies.
That second amendment, however, drew some uncertainty from departmental experts, who told MPs it remained unclear whether the deadline would actually speed up removals compared to current practice.
Sexual deepfakes are one piece of a much larger bill. Bill C-16 also takes aim at coercive control and seeks to restore mandatory minimum prison sentences that courts had previously struck down as unconstitutional. It sits within a trio of tougher-on-crime measures introduced by the Liberal government, none of which have yet cleared Parliament.
The bill now moves to the Senate, where it must pass before it can become law.
