In early 2020, deepfake expert Henry Ajder spotlighted a troubling advancement in technology—a Telegram bot capable of "undressing" photos of women, using artificial intelligence. The bot had already been used to generate over 100,000 explicit images by the time it came to attention, a distressing number that included images of children. Ajder described this development as a "watershed" moment in the potential for deepfake technology to cause harm.
Over the subsequent years, the proliferation of deepfake technology has only intensified, becoming more accessible and widespread. A recent review undertaken by WIRED revealed at least 50 such bots hosted on Telegram, a messaging platform, which claim to create explicit photos or videos of individuals through a few simple clicks. These bots demonstrate varying capacities, with some advertising the ability to "remove clothes" from images, while others purport to generate images of individuals engaged in various sexual activities.
The scale of their usage is notable. The reviewed bots boast a cumulative user base of over 4 million monthly users, according to data presented by the individual bots themselves. Two of these bots claim in excess of 400,000 monthly users each, whereas another 14 report having more than 100,000 users each. This illustrates not just the extent of explicit deepfake creation tools, but also cements Telegram’s status as a significant hub for such activities. However, the review primarily focused on English-language bots, indicating that this may only be a fraction of the existing scene on Telegram.
Ajder expressed deep concern over the ease of access to these tools on a major global app, particularly noting the severe impact on women and young girls. "It is really concerning that these tools—which are really ruining lives and creating a very nightmarish scenario primarily for young girls and for women—are still so easy to access and to find on the surface web, on one of the biggest apps in the world," he stated.
Initially appearing towards the end of 2017, explicit nonconsensual deepfake content, categorised as nonconsensual intimate image abuse (NCII), has since burgeoned. This growth has been significantly propelled by advancements in generative artificial intelligence. Across the digital landscape, an array of "nudify" and "undress" websites, alongside more refined tools and Telegram bots, have emerged—targeting thousands of women and girls worldwide, from high-profile figures like Italy's prime minister to schoolgirls in South Korea. A recent survey startlingly revealed that 40 percent of US students had encountered deepfakes related to their K-12 schools in the past year.
The identified Telegram bots are bolstered by at least 25 associated Telegram channels, which collectively engage more than 3 million members. These channels serve as platforms for users to receive updates about new bot features and special offers on "tokens" required to operate the bots. Furthermore, they are often essential in guiding users to alternative bots when an existing one is taken down by Telegram.
As this phenomenon continues to expand, it highlights the unprecedented challenges posed by technological advancements in managing privacy and consent in the digital era, particularly concerning deepfakes and their misuse on widely accessible platforms like Telegram.
Source: Noah Wire Services