YouTube and X are sending large amounts of traffic to websites that create nonconsensual explicit deepfakes, according to a new report from the Institute for Strategic Dialogue. That finding is ugly for victims and awkward for platforms that already say their rules ban this stuff.
The ISD, an organization that researches extremism and disinformation, examined the 10 most-used apps and websites for making so-called nudify images, along with the routes people took to find them. These tools generate synthetic nude or sexualized images of people without their permission, often from ordinary photos.
According to the report, social networks referred more than 5.7 million visits to nudify sites from December 2025 through March 2026. YouTube was the largest source, accounting for 1.82 million visits, or more than 30 percent of the social referral traffic. X ranked second, with more than 1.3 million visits.
The mechanics are not mysterious. ISD said videos surfaced through searches for terms such as “undress app” and “nudify app.” Some videos reviewed or promoted particular services. Others pointed viewers to discount codes or free credits. In other words, the platforms were not merely hosting discussion of the tools; in some cases, they were helping users get to the checkout line.
Melanie Smith, ISD’s senior director of research and policy, told WIRED that YouTube was not just a passive source of referral traffic. She said many examples helped people use the tools. Smith also told WIRED that YouTube’s rules bar links to or ads for sexually explicit sites, which should cover nonconsensual imagery, revenge porn and nude leaks in theory, but did not appear to be enforced consistently.
The ISD report said the material it found seemed to conflict with YouTube’s ban on sexually explicit content. The report argued that such a ban should include sites or services built to generate nonconsensual explicit images, yet that content was still easy to find on the platform.
YouTube spokesperson Boot Bullwinkle told WIRED the company has strict rules against “unwanted sexualization,” including nonconsensually shared intimate imagery. He said those policies apply both to YouTube content and external links, and include altered or synthetic material that realistically depicts nudity.
X did not respond to WIRED’s request for comment. In January 2026, X faced criticism after users generated nude or sexually suggestive images of women, including some minors, with the Grok chatbot. The company later said it was limiting Grok access to paying users and had zero tolerance for child sexual exploitation, nonconsensual nudity and unwanted sexual content.
ISD also found that some nudify services charged as little as $1 per image. WIRED previously reported that such sites may collectively be bringing in up to $36 million in revenue.
The report said common targets included current and former girlfriends, as well as relatives including sisters and cousins. Smith told WIRED that some users’ aims were not sexual, citing requests intended to get people fired or otherwise damage their lives.
Nonconsensual intimate imagery is illegal in the United States. The federal Take It Down Act, fully in effect since May, requires social platforms to remove nonconsensually distributed images within 48 hours after a victim submits a takedown request. Most states have passed some form of anti-deepfake law, and Minnesota became the first state to specifically ban nudification apps in May 2026.
ISD called for coordinated responses across platforms, policy and education, including stronger platform regulation and funding for school digital literacy programs. WIRED has reported deepfake cases in more than 90 schools worldwide, which gives the policy debate a grimly practical clock.
This story draws on original reporting from WIRED.
