Telegram has become a hub for explicit, nonconsensual deepfakes, with over 50 bots generating fake nude images and videos of women and minors, reaching more than 4 million users each month. One such bot was first discovered in 2020, producing over 100,000 explicit images, including those of children.
Despite recent removals, the ease of creating such harmful content raises concerns about Telegram’s lax enforcement and the growing impact of AI-powered abuse. Since then, the problem has escalated, with bots now becoming a common tool for generating explicit deepfakes on a larger scale, largely enabled by Telegram.
A recent investigation by Wired has revealed a thriving ecosystem of at least 50 Telegram bots, each claiming to generate explicit photos or videos with just a few clicks. These bots boast a combined user base of over 4 million monthly users.
While some bots personalise in ‘removing clothes’ from photos, others purport to create even more graphic content. The findings illustrate how far the technology has advanced and how Telegram has become a prominent platform for distributing nonconsensual deepfake content.
Wired’s review of Telegram revealed 25 additional channels, with over three million subscribers, designed to promote these deepfake bots. These channels often update users about new features and offer tokens for purchase, allowing users to access the bots.
Although Telegram removed 75 bots and channels following Wired’s inquiry, the problem persists, with new bots quickly surfacing to replace those taken down.
The ease of bot creation on Telegram, combined with the app’s search functionality and bot-hosting capabilities, makes it a fertile ground for deepfake abuse. Bots can be easily located, shared, and operated with little oversight. While some bots require users to accept terms of service prohibiting uploading nonconsensual images, enforcement of these rules appears nonexistent.
The ecosystem around deepfake generation has become so lucrative that some cybercriminals are now creating fake ‘nudify’ websites to distribute malware.
Telegram has long been criticised for its hands-off approach to content moderation. Despite removing some harmful bots and channels, the platform has been slow to take meaningful action against deepfake creators.
In August 2024, Telegram CEO Pavel Durov was arrested in France as a part of an investigation into CSAM. Just a month after, in September, South Korea also initiated a probe into Telegram’s alleged negligence in addressing deepfake-related sex crimes.
In the News: WhatsApp bans 8.4 million Indian scam accounts
