Who Is Mr. Deepfakes? Paris Hilton's TikTok Series Reveals The 'Nice Guy' Behind The Fallen AI Porn Empire

A new investigative series is shifting the focus from anonymous algorithms to the individuals exploiting them. The Searching for Mr. Deepfakes documentary series, a 13-part documentary series released directly on TikTok and launched by Paris Hilton 11:11 Media in collaboration with journalist Laurie Segall, pulls the curtain back on one of the most notorious hubs for AI deepfake pornography.

The documentary highlights a jarring reality: the primary figure allegedly controlling the platform was not a shadowy international cybercriminal, but a suburban pharmacist who led a seemingly ordinary life.

The site, which was shuttered in May 2025 after a sustained investigation, served as a grim marketplace where millions of videos featuring non-consensual imagery were hosted and distributed. The series follows a three-year investigation into the anonymous operator of MrDeepFakes, a site Segall described as 'one of the most dystopian websites I'd ever seen', which at its peak drew 17 million monthly visitors before being forced offline in 2025.

The contrast at the heart of the story is jarring. According to a joint investigation by Bellingcat, CBC News, and the Danish publications Politiken and Tjekdet, the most prominent figure identified as controlling MrDeepFakes was not a shadowy hacker or an international crime lord. He was a hospital pharmacist from the Toronto suburbs who drove a Tesla and posted family photos online.

The Man Behind the Username

David Do, a 36-year-old pharmacist working within the Oak Valley Health hospital network in the Greater Toronto Area, was identified by Bellingcat and its partners as having played a prominent role in the administration of MrDeepFakes.

Investigators pieced together his alleged involvement using open-source data, cross-referencing credential leaks, IP addresses, repeated usernames, and a unique password to construct a more-than-decade-long digital trail linking him to the site.

Do was not merely a passive administrator; he reportedly produced his own deepfake content and actively assisted community members in generating their own material. In archived posts, a user linked to Do declared he was 'dedicated' to improving the platform, adding: 'There is a reason why we are the biggest deepfake site. I care about the community and teaching others.'

The issue sits within a wider pattern of abuse linked to AI-generated sexual content. 'Paris Hilton was one of the earliest victims of this type of abuse, and now, fast-forward all these years later, she's one of the earliest victims of deepfake pornography,' Segall said. '[N]one of us are safe. In this AI-driven era, to become a victim, you don't have to take a nude picture. You just have to exist.'

When CBC's visual investigations unit hand-delivered correspondence outlining the findings to Do in April 2025, he did not respond. Shortly afterwards, his Facebook page and some family members' social media accounts were deleted. He then travelled to Portugal with his family, only returning to Canada the week the story broke. When confronted by a CBC journalist, he said: 'I don't want to be recorded, please. I have to go. I'm busy right now.'

Oak Valley Health placed Do on leave immediately following publication of the investigation. By 15 May 2025, he was no longer an employee of the network. The Ontario College of Pharmacists separately announced it was taking steps to investigate and determine what action was necessary to protect the public. He has not publicly commented on the allegations against him.

Inside the Platform

MrDeepFakes, which operated from 2018 until its forced closure in May 2025, billed itself as 'the largest and most user-friendly' platform for celebrity deepfake pornography. At its peak, it hosted close to 70,000 explicit and sometimes violent videos, which had collectively been viewed more than 2.2 billion times. A community of more than 650,000 members operated in the background, sharing tips, commissioning custom content, and posting derogatory commentary about their targets.

Research shows that 99 per cent of deepfake pornography victims are women. The site's targets ranged from Hollywood actresses and pop stars to politicians, YouTubers, streamers, and private individuals who had never sought public attention. In 2020, the deepfake detection company Sensity reported that 96 per cent of all detected deepfakes were sexualised in nature, with women the overwhelming majority of targets.

Adam Dodge, founder of EndTAB (End Technology-Enabled Abuse), described MrDeepFakes as an 'early adopter' of deepfake technology that had evolved from a video-sharing platform into a training ground and marketplace for AI-powered sexual abuse material of both celebrities and private individuals.

'Our digital world is really good at empowering people who want to do harm by allowing them to remain anonymous while simultaneously making it almost impossible for victims to unmask them,' he said.

The site had already begun blocking traffic from the United Kingdom in April 2024 in anticipation of proposed legislation targeting the non-consensual production of deepfake material. Its closure in the Netherlands and UK preceded its global disappearance, following the CBC-led investigation, and a week after the US Congress passed the Take It Down Act, which criminalises the distribution of non-consensual deepfake pornography.

The Women Whose Lives Were Upended

The statistics can obscure what each data point represents: a real woman who did not consent to the creation or distribution of sexualised imagery of herself, and who often found it impossible to put the genie back in the bottle once that content spread online.

German reporter Patrizia Schlosser is one of many such women. She tracked down the man who had uploaded more than 30 non-consensual sexually explicit images of her to MrDeepFakes, which required a determined investigation of her own, and one that highlighted how isolated victims can become in the absence of robust legal recourse.

'It was just one of the most painful and traumatising, humiliating, degrading experiences of my life', Hilton said, speaking about her own experience of non-consensual content, in Searching for Mr. Deepfakes

Research published in peer-reviewed journals documents the scale of harm deepfake abuse inflicts: victims frequently experience severe emotional distress, including depression, anxiety and post-traumatic stress disorder, alongside economic instability, reputational damage, and social isolation.

The removal burden placed on survivors is described as cruelly unrealistic: they are expected to track down every copy of non-consensual imagery across multiple platforms and repeatedly report each instance, with no guarantee of timely or effective removal.

In one 2025 UK case, journalist Daisy Dixon discovered AI-generated sexualised images of herself on X, created using the platform's own Grok AI tool; it took days for the platform to geoblock the function while the content continued to spread.

Why Paris Hilton — and Why TikTok

The choice of Hilton as a co-producer is deliberate and deeply personal. Hilton was among the first celebrities targeted by deepfake pornography, which marks a distinction she holds in addition to having been an early victim of non-consensual intimate image distribution in the early 2000s.

Hilton has also translated that personal history into political action. In January 2026, she joined Congresswoman Alexandria Ocasio-Cortez on Capitol Hill to advocate for legal protections for deepfake abuse victims, calling it 'the newest form of victimisation.'

The decision to release Searching for Mr. Deepfakes on TikTok rather than through a traditional broadcaster or streaming platform was strategic. Segall, a former CNN editor-at-large who spent a decade covering technology before founding Mostly Human Media, said she chose TikTok because it 'lends itself to the format' of 'rabbit hole' storytelling, and because it places the content directly in front of younger women — the demographic most at risk.

'We really want people who need this the most to see it,' Segall has said. Choosing TikTok is also an act of provocation: the platform where much of this abuse spreads and where victims first discover that fake content about them exists becomes the platform where the story of accountability is told.

The Law Catches Up Slowly

Legal frameworks have struggled to keep pace with the technology. Segall noted that this pattern, where tech moves quickly while laws lag, with the earliest victims disproportionately women, was familiar from earlier waves of online abuse, including non-consensual pornography.

In the UK, the Online Safety Act 2023 included provisions covering non-consensual intimate imagery, but campaigners argue enforcement remains inconsistent and that the act does not go far enough in holding platforms to account for hosting deepfake material. Further legislation aimed at extending police powers in cases of intimate image abuse was at its second reading in the House of Lords as of late 2025.

In the United States, the passage of the Take It Down Act in April 2025 represented a meaningful step, but critics note that criminalising distribution does not address how easily the content is created in the first place, nor does it prevent it from migrating to jurisdictions with weaker protections.

Key Facts & Figures

• MrDeepFakes operated for seven years, from 2018 to May 2025, hosting nearly 70,000 explicit videos that were viewed more than 2.2 billion times.
• At its peak, the site attracted 17 million monthly visitors.
• Research shows 99 per cent of deepfake pornography victims are women.
• More than half of deepfake abuse victims in the US have contemplated suicide, according to UN Women.
• David Do was terminated from Oak Valley Health on 15 May 2025. He has not publicly commented on the allegations.
• The US Take It Down Act, criminalising the distribution of non-consensual deepfake pornography, was passed by Congress in April 2025.

The broader industry context makes the regulatory challenge clear. A 2024 survey found that at least one in nine high school students in the US knew of someone who had used AI to make deepfake pornography of a classmate. The technology is no longer the preserve of technical specialists — free or cheap applications make it accessible to virtually anyone. In that environment, even the closure of the web's most notorious marketplace does not remove the threat; it simply disperses it. That is the final message at the heart of Searching for Mr. Deepfakes. The site is gone. The man allegedly at its centre has lost his job and his anonymity. But the tools that made it possible remain widely available, the laws that might constrain their misuse remain incomplete, and the women who were harmed are still living with the consequences.

As Segall put it: 'When it comes to the rise of sexually explicit deepfakes, we've seen a version of this story before: the one where tech moves quickly, the laws haven't caught up, and the earliest victims are women.'

The technology is outpacing the law, and until platforms are held strictly liable for the material they host, the threat will likely simply disperse to other jurisdictions. The story of MrDeepFakes serves as a final warning: while the platform is gone, the tools for abuse remain in the hands of anyone with an internet connection.