Blushing because AI 18+ movies are spreading on social networks
(Dan Tri) - Videos created by AI with sensitive images and details suddenly appear while surfing social networks, causing many users to fall into awkward situations.
Afraid to open the phone in crowded places
In the evening, as usual, Tuan (23 years old, Hai Phong) opens his phone to surf social networks for entertainment after a day of work. Between regular videos, he suddenly came across an advertisement for a movie created with artificial intelligence (AI).
The video uses curious titles and images of sexy female characters. Attracted by the opening, Tuan stopped to watch for a while longer. However, when the content gradually became sensitive and inappropriate, he decided to skip it.
Tuan thought this was just a random ad. But in the following days, videos promoting AI movies with similar content continuously appeared on phone screens.
He said that even when switching to another platform, he still sometimes encounters images of similar nature.
Tuan said he had many times fallen into awkward situations when AI movie advertisements with sensitive content suddenly appeared while surfing social networks (Photo: Doan Thuy).
One time, when Tuan was sitting in a crowded place, an AI movie advertisement with sensitive images suddenly appeared. The phone's sound was not turned off at that time, causing some people around to turn and look.
"I just had time to turn off my phone. At that time, I was very embarrassed because people around me might think I was actively watching such content," Tuan said.
Not only Tuan, some users that Dan Tri reporter spoke to also said they had fallen into similar situations.
Phuong (23 years old, Quang Ninh) often watches short videos created with AI. Content with dramatic details and curious titles often make her stop longer to see what happens next.
One time, when the commercial ended at an interesting plot point, Phuong downloaded the introduced application to continue watching.
According to her, initially the films mainly revolved around love stories, family stories or dramatic details. After a while, Phuong noticed that ads with 18+ elements appeared more often.
Phuong noticed that AI movie ads with 18+ elements appear more often after a period of watching and interacting with AI movie content on social networks (Screenshot).
Some videos only insert a few seconds of sexual images at the beginning and then move on to introducing the main content of the movie.
"I was sitting next to an acquaintance and was about to show them a funny video I just saw. Just as I scrolled to the next video, a sensitive ad appeared. I had to swipe away immediately, both uncomfortable and embarrassed," Phuong said.
From a video to a "matrix" of AI 18+ content
From the ads appearing on social networks, Dan Tri reporters tried accessing some of the attached links.
After just a few taps, viewers are redirected from social networks to applications that provide a series of short films created by AI with many different topics, including many content with sexual elements.
Notably, during the experience, the reporter noted that some services only require users to create or log in to an account to continue viewing, without the step of verifying the actual age of the visitor.
This means that a link from an ad appearing publicly on a social network can take users to sensitive content without encountering an obvious age-check "fence".
Some ads not only introduce AI movies but also lead users to "AI girlfriend" creation services, allowing character customization and sexual content creation on demand (Screenshot).
Some advertisements are even more daring when they do not stop at introducing AI 18+ movies. Viewers are invited to access services that allow them to create their own "AI girlfriends", customize characters according to their preferences and create sexual content using AI.
The situation of advertising AI services with sexual content is not only recognized by the media in Vietnam.
In April 2024, WIRED surveyed Meta's ad library and discovered at least 29,000 ads had been posted to promote "AI girlfriend" applications with adult elements.
Of these, at least 19,000 ads use the term "NSFW" (short for "Not Safe For Work", a term often used to refer to sensitive content that is not suitable for viewing at work or in public places) and 14,000 ads tout the service "NSFW AI". Some ads lead users to chatbots that provide AI-generated sexual images and conversations.
By June 2025, CBS News continued to detect at least hundreds of ads on the Meta ecosystem promoting "nudify" tools - using AI to create fake nude images from photos of real people.
Investigations by WIRED and CBS News have recorded a large number of ads related to “AI girlfriend”, NSFW and the “nudify” tool appearing on the Meta ecosystem (Graphic: AI)
Some links in the ads take users to websites that provide additional features to create sexual content using AI.
Meta then said it removed the detected ads, removed the pages behind them, and blocked related URLs.
Notably, CBS News also examined a "nudify" website (using AI to create fake nude images from real people's photos) promoted on Instagram and noted that this site does not require age verification before users upload photos.
Advertisements of this type continued to be a problem in the years that followed. Meta once admitted to CBS News that the entities behind applications that exploit AI to create sensitive content constantly change their methods to avoid detection systems.
The company said it had to develop additional technology to identify ads that use seemingly harmless images in order to pass the review process.
For users like Tuan and Phuong, the question is not just why sensitive ads can appear publicly on social networks. What makes it more difficult for them to understand is why after a few times of encountering or stopping to watch, the same content seems to appear more and more on the screen.
Why does this happen and how do social network algorithms really "understand" user behavior? Dan Tri will update in the next article.