“We’ve been trying to reach you regarding MrDeepFakes,” the journalist said, handing Do a list of findings. “I want to ensure you’ve received our emails.”
“I don’t know anything about that,” Do replied. “I’m at work right now.”
Then he shut his office door.
Until May 2025, it was the world’s largest and most active marketplace for AI-generated pornographic “deepfakes.”
To colleagues at Oak Valley Health, Do was a mild-mannered pharmacist who earned $121,000 a year ensuring patients at the network’s two hospitals got the right pills. He drove a Tesla, lived in a comfortable home outside Toronto with his partner and, according to hospital PR, was part of a team “ensuring safe practices.”
But online, Do was “dpfks”—administrator, forum czar, video creator and technical mastermind behind MrDeepFakes.com, a now-defunct website for sharing nonconsensual AI-generated sexual and violent videos featuring the faces of famous women. Until May 2025, it was the world’s largest and most active marketplace for AI-generated pornographic “deepfakes.”
MrDeepFakes.com launched under the radar in 2018 and ballooned to more than 650,000 registered users, 70,000 videos and a jaw-dropping 2.2 billion views. It offered everything from celebrity face-swaps to violent fantasies featuring politicians, social media influencers and private citizens—almost all of them women.
On its forums, customers could pay creators to digitally strip, humiliate and sexually assault anyone—from Scarlett Johansson to a user’s unsuspecting girlfriend. Payment was in crypto. The damage—incalculable.
The story of MrDeepFakes isn’t just about a pharmacist with a secret. It’s a global warning shot about the unchecked rise of AI-fueled sexual violence—and the yawning gap between technological harm and legal accountability. In Canada, where Do operated with impunity, no criminal statute explicitly outlaws nonconsensual adult deepfake porn. And while the US, Britain, Denmark and other countries scramble to legislate against this digital plague, the victims—disproportionately women—are left to navigate a justice system not built for their era. With Do unmasked and MrDeepFakes shuttered, the real reckoning has only just begun.
On May 4, following mounting pressure from the Canadian Broadcasting Corporation (CBC), a Danish newspaper called Politiken, the investigative journalism outlet Bellingcat, and a Danish fact-checking outfit named TjekDet, MrDeepFakes vanished. In its place: a somber black screen.
“A critical service provider has terminated service permanently,” a message on the site’s home page stated, adding: “We will not be relaunching. Any website claiming this is fake.”
By then, investigators had amassed a sprawling digital dossier tying Do to the site’s inner workings—from email trails and burner accounts to leaked passwords and photos, many linked to a Gmail account that used the characters “dpfks,” which once sat quietly in the site’s source code. That string of five consonants and a single forgotten email became the loose thread that unraveled a multimillion-view empire of abuse.
And when a CBC reporter confronted Do again outside his Tesla, the pharmacist had little to say.
“I don’t want to be recorded,” he said. “I have to go.”
And then, just like that, he did.
Do’s descent into porn site administration didn’t begin with deepfakes. In 2008, as a student at the University of Waterloo, he ran a warez forum—an online hub for sharing pirated digital content like software, games or movies, illegally copied and distributed without permission, often carrying legal and security risks—employing a username that also appeared across databases tied to porn marketing.
Tracking down Do was a global feat of forensic journalism.
By 2018, just as Reddit banned its deepfake porn community, a new site appeared: dpfks.com. It soon became MrDeepFakes, and “dpfks” was its most active and visible administrator.
“I just got home from my day job,” dpfks wrote in 2018. “Now back to this!”
The “this” was a growing empire. On forums, dpfks posted rules, offered tutorials and commissioned moderators. He boasted about the site’s dominance and shared hundreds of videos himself—at least 161 deepfakes under his handle.
Some were violent. Others featured politicians like Alexandria Ocasio-Cortez and climate activist Greta Thunberg in sexually degrading scenes. One post included a dataset of over 6,000 images of Ocasio-Cortez to aid other users in creating new fakes.
“[It’s] quite violating,” said Vancouver-based YouTuber Sarah Z., who discovered her own likeness had been deepfaked on the site. “These are real people … who often suffer reputational and psychological damage.”
MrDeepFakes wasn’t just a community—it was a business. Users paid hundreds of dollars for bespoke videos. In 2020, Do—under the handle “dj01039” posted that he was making up to $7,000 a month. By 2025, the site registered over 18 million visits every four weeks. Advertisers lined up. New creators flocked in. And the targets—nearly all women—had no say in the matter.
The forum’s ethos was summed up by a tagline posted by dpfks himself: “Fake it till you make it.”
But behind that glib slogan was a darker logic—one that, according to researcher Penille Rasmussen, caters to “gaining power over women. Being able to control her or do something against her will.”
That control wasn’t just digital. It became structural.
When a Danish investigator searched for male politicians on MrDeepFakes, she found none. But their wives and daughters—and in one case, even a granddaughter? “Fair game.”
Conservative MP Mai Mercado, whose face appeared in a manipulated pornographic image, said the experience was tantamount to a death threat. “It’s on the same level as a verbal threat of being killed,” as she put it.
As the tech evolved, the law lagged. In Denmark, sharing nonconsensual pornography is illegal. The government has also introduced legislation to criminalize creating it in the first place. In Canada? Still legal, unless a minor is involved.
Prime Minister Mark Carney, amid growing outrage, pledged to criminalize deepfake porn. “We will make producing and distributing nonconsensual sexual deepfakes a criminal offense,” he said during his campaign.
But legislation remains pending.
Lawyer Miriam Michaelsen, who represents Danish victims, said the legal vacuum is devastating. “When it’s a foreign platform and perpetrator, it’s currently a virtually consequence-free space because authorities lack both the resources and capabilities to handle these cases and support victims,” she explained, adding: “Many victims don’t have a chance to remove images and videos.”
Even now, Do has not been charged. But Denmark is weighing an extradition request under its nonconsensual image abuse laws, which offer broader protections.
In a world where technology outruns ethics, even those sworn to do no harm can leave behind a quiet trail of devastation.
The faces on MrDeepFakes weren’t just those of Hollywood stars. They belonged to athletes, environmental activists, Twitch streamers and everyday women with Instagram accounts. One of the site’s rules stated that only influencers with more than 120,000 followers could be faked without consent—but CBC found content featuring people with far fewer.
Some users requested fakes of their wives. Others traded tips on the messaging app Telegram. One wrote bluntly: “Just pure anal for her.”
For law professor Suzie Dunn, the implications are chilling.
“They see the women in these images as digital objects,” she said. “It really discourages people from going into politics … even being a celebrity.” It discourages something deeper, too: the basic right to exist publicly as a woman without fear of digital violation.
For years, the power of platforms like MrDeepFakes rested on one principle: anonymity. But the Do investigation—spanning Canada, Denmark and the Netherlands—proved that even the most meticulous secrecy can unravel.
Do’s password was reused. His burner accounts linked back to his name. A photo on an Airbnb profile matched the face in Oak Valley Health’s staff picture. His fingerprints were all over the code.
Tracking down Do was a global feat of forensic journalism. Netherlands-based Bellingcat traced the domain. CBC tracked the Tesla. Politiken and TjekDet followed the usernames.
And once Do was named, the silence that had shielded perpetrators, shut down victims and lulled institutions into inaction finally gave way.
MrDeepFakes is gone. But its videos live on in backup drives, re-uploads and dark web forums. And a new platform is surely waiting to rise in its place.
For Sarah Z., for Queen Mary of Denmark, for the many who never even knew their faces were used—justice remains elusive.
“There’s only so much that I as an individual can do,” Sarah Z. said. “Any change will have to be legislative and systemic.”
The man who once promised to “keep this [deepfake] community running as long as I can” now says nothing. His site is dead. His aliases erased. His future uncertain.
Even so, his digital fingerprints endure—as a blueprint for abuse, and as a cautionary tale for a world unprepared for the weaponization of AI.
On May 13, Oak Valley Health confirmed that Do had been placed on leave from his pharmacy post and that an internal investigation into his alleged crimes was underway—the white walls of medicine no longer shielding the shadow he cast online. The institutional pause may have come too late for the countless women whose likenesses were exploited while Do quietly played healer by day and digital predator by night.
His case underscores a grim truth: in a world where technology outruns ethics, even those sworn to do no harm can leave behind a quiet trail of devastation. In the end, the pharmacist who once promised to ensure “safe practices” helped build one of the most dangerous platforms of the internet age.
And when the reckoning came, he did exactly what his avatars always could:
He disappeared.
