YouTube Drives 1.8M Visits to Nonconsensual Deepfake Sites

Mainstream social media platforms are serving as major gateways to websites and apps that generate nonconsensual explicit deepfakes, despite policies explicitly prohibiting such content, according to new research.

The Institute for Strategic Dialogue, an organization focused on combating extremism and disinformation, analyzed the top 10 nudification platforms and traced how users discover them. The findings reveal that social networks drove more than 5.7 million visits to these sites between December 2025 and March 2026, with YouTube and X accounting for the vast majority of referral traffic.

YouTube leads referral traffic despite policy prohibitions

YouTube was responsible for 1.82 million visits to nudify sites—over 30 percent of all social media referrals tracked in the study. These visits originated from videos promoting specific apps, providing review content, and sharing promotional codes for free credits. Users found this content through straightforward searches for terms like "undress app" and "nudify app."

X ranked second, accounting for more than 1.3 million visits during the same period.

The research team noted these patterns directly contradict YouTube's stated policies, which prohibit sexually explicit content and ban links to or advertisements for sexually explicit websites. YouTube spokesperson Boot Bullwinkle told WIRED the platform maintains strict policies against unwanted sexualization and nonconsensually shared intimate imagery, including synthetic content that simulates nudity. These rules apply to both on-platform content and external links.

Why it matters

The scale of mainstream platform involvement in directing users to nudification tools undermines efforts to combat nonconsensual intimate imagery. With these apps generating an estimated $36 million in collective annual revenue while costing users as little as $1 per image, the business model thrives on accessibility. The research found victims include current and former romantic partners, family members including sisters and cousins, and targets selected not for sexual purposes but to damage careers and reputations.

Enforcement gaps persist despite new legislation

The federal Take It Down Act, which took full effect in May 2025, requires social media platforms to remove nonconsensually distributed images within 48 hours of a victim's takedown request. Minnesota became the first state to specifically ban nudification apps in May 2026, and most states have adopted some form of anti-deepfake legislation.

Yet the tools continue proliferating. The ISD study emphasizes that effective responses require coordinated action across multiple domains: platform regulation, policy interventions, and educational initiatives including digital literacy programs in schools. A separate WIRED investigation documented reported deepfake cases in more than 90 schools worldwide.

Melanie Smith, senior director of research and policy at ISD, emphasized that YouTube's role extends beyond passive referral traffic. The platform actively facilitates use of these tools through hosted promotional content, she noted.

X faced significant criticism in January 2026 when users exploited its AI chatbot Grok to generate nonconsensual nude or sexually suggestive images, including some depicting minors. The company subsequently limited Grok access to paid users and issued a statement reaffirming zero tolerance for child sexual exploitation, nonconsensual nudity, and unwanted sexual content. X did not respond to requests for comment on the new research.

These findings were first reported by WIRED, based on the Institute for Strategic Dialogue study published Monday.