Have you ever seen a bikini-clad reporter strolling in a busy market and indulging in vulgar & sexually-explicit conversation with random people? Or a female scantily-dressed podcaster discussing sexual activities and dark fantasies with a grandma? Or a teacher exchanging sexual jokes with minor students?
Sounds unfamiliar? Not anymore. Content creators are now exploiting Artificial Intelligence for generating hyper-realistic erotica videos to attract social media engagement.
A review by India Today’s Open-Source Intelligence (OSINT) team found such AI content is widely popular on Instagram and YouTube, producing engagement statistics in such a short span that would seem like a dream to most creators.
Videos of this nature were billed as “vulgar” and “obscene” and at least 43 OTT apps distributing such content were banned, the Ministry of Information & Broadcasting told parliament earlier this year.
We reviewed more than two dozen Instagram accounts and YouTube channels dishing out AI-generated racy videos on scale. The content of these videos include sexual banter, chats and jokes about sexual activities, and private parts.
The gap between the AI avatars featured in these videos are very stark, often featuring a person in their 20s and the other in their 70s or 80s. Even schoolgirls and toddlers have not been spared. In many videos, minor children are made to utter sexually-charged lines and jokes.
VULGARITY FOR VIEWS
In many cases, the content stretches into long explicit sequences involving nudity. Some creators go further, pushing out highly graphic AI videos that simulate intimate encounters in public places, feeding a market of increasingly extreme fantasies.
Beyond avatars, creators generate explicit AI images resembling popular Bollywood actresses.
On Instagram, one such account is AI Wi**. It has over 582,000 followers. Its feed showcases hyper-sexualised AI explicit images of women in skimpy attire, cast in roles instantly recognisable to Indian audiences, washing clothes, sweeping floors, teaching in classrooms, cooking in kitchens, even working on construction sites.
These accounts draw lakhs of followers and subscribers and rake up crores of views within months. A YouTube channel created just two months ago, on August 3, 2025, now boasts more than 12 crore views. Some ‘Shorts’ featuring AI avatars on the channel have gained more than 2 crores views each.
MILKING THE AI TECH
Such videos are not just for fun and social media engagement, they are used to generate profit for their creators. Accounts publishing these videos offer premium content for subscription, advertise marketing tips imparted over 15-minute video calls, and even provide courses on how to create similar AI videos.
AI profiles like “Ka**ka Sharma” advertise 15-minute video calls for 499 while “Z*** Wali Didi” sells courses to curious users who see an opportunity in generative AI technology. Some accounts provide access to closed online groups for a fixed fee.
Creators use advanced freemium AI tools to produce these realistic videos. Many of these videos carry the watermark of Google Veo, a text-to-video model launched last year that can generate hyper-realistic visuals. Some others were created by Tensor Art, an online platform that supports image-to-image, text-to-image, and image-to-video generation, an admin in a closed Telegram group informed its 1300-odd members.
CONCERNS OF REAL-WORLD HARM
Researchers say that overexposure to sexualised content – AI generated or not – can lead to several behavioral and psychological issues as well as distorted belief about sex among internet users, particularly younger people.
A 2024 study by Northeastern University found that Instagram filled kids’ feeds with sexually suggestive content, including videos from adult content creators. The investigation was done by creating Instagram accounts with the age set at 13.
The distorted portrayal of sex in such videos can lead to premature risky sexual behavior, anxiety, depression, and body dissatisfaction. Younger people may also develop unrealistic and unhealthy attitudes toward intimacy, consent, and sexuality.
WHAT THE LAW SAYS
In a 2024 press release, the Ministry of Information and Broadcasting said that it had banned 18 OTT platforms for “obscene and vulgar content” under provisions of the IT Act, the Indecent Representation of Women (Prohibition) Act, 1986, and the now-defunct IPC.
But there remain loopholes in the regulation of obscene content on the Indian internet. In April, the Supreme Court said legislative measures were necessary to address the menace of online obscenity.
There are no laws to regulate AI-generated obscenity in India. Earlier this month, the I&B Minister Ashwini Vaishnav said that the government was mulling a new law to regulate deepfake content.
“At the moment, beyond a patchwork of laws like the Bhartiya Nyaya Samhita and the Information Technology Act, 2000, India has no specific legal instruments to regulate explicit AI-generated content. Rule 3(1)(b) of the 2021 IT Rules requires platforms to prevent obscene or pornographic content and protect individual privacy, and explicit AI material falls within this ambit,” says Abhivardhan, founder of Indic Pacific Legal Research.
Social media platforms do have their policies on obscenity and AI-generated content, but they seem unable to keep up.
Meta says its policies restrict sexually explicit language to users under the age of 18 and ban display of adult nudity or sexual activity. YouTube also prohibits explicit content created for sexual gratification. Yet enforcement is inconsistent, and the flood of AI explicit content shows how rules often lag behind the reality of the feeds.
– Ends