50 AI Bots With Over 4 Million Users Used To Create Deepfakes Of Women And Girls On Telegram: Report

Four years after a deepfake expert, Henry Ajder, discovered that a Telegram bot had been designed to “undress” pictures of women using artificial intelligence, the menace is bigger than ever. A new study has now found that at least 50 such bots are now active on the platform and they have over 4 million “monthly users” combined.

These tools are being used to generate nude images of real-life people. Some of these bots are programmed to edit photographs within a few clicks, assisting users to generate images that strip clothes or even show people performing sexual acts.

According to WIRED’s analysis of Telegram communities that engage in explicit content, at least two bots had over 400,000 monthly users, while another 14 had more than 100,000 subscribers. Ajder has labelled this situation as “nightmarish” while highlighting the harm these tools present to people, especially girls.

These bots are mainly used to exploit young girls and women, leading to sextortion. According to a recent survey report by the Center for Democracy and Technology, as many as 40% of the students in the United States claimed that they were exposed to deepfakes in their learning institutions. The accessibility of such tools has risen, meaning that more people get to misuse such tools thereby causing whoever is on the other end of the tool extreme discomfort and distress.

Creating and sharing non consensual intimate images (NCIIs) can cause significant psychological harm to victims. Emma Pickering from the UK’s largest domestic abuse organization, Refuge, explained to WIRED that the said images trigger feelings such as humiliation, fear and shame that greatly affect mental well-being. This situation is made worse by the fact that there are no repercussions for the perpetrators.

Even though there have been continued debates and discussions as to how to control such technologies, the precise means of managing them have not been seen. Some legislation such as the Deepfake Accountability Act in the United States has been proposed, but these measures seem to have not yet effectively addressed the issue of restraining the use of these bots.

The reaction of Telegram has been described as inadequate. While some of the bots were deleted after media coverage, new ones appeared almost immediately.