Abusive AI image tools are no longer hidden at the edge of the internet. A WIRED review of Telegram communities found a large ecosystem of bots that claim to create explicit images or videos from user-submitted photos, often with only a few clicks.
The review identified at least 50 bots connected to explicit nonconsensual content. Together, those bots listed more than 4 million “monthly users,” according to the statistics shown by each bot.
A sharp expansion from early deepfake abuse
In early 2020, deepfake expert Henry Ajder uncovered one of the first Telegram bots built to “undress” photos of women using artificial intelligence. He recalls that the bot had been used to generate more than 100,000 explicit photos, including those of children.
Ajder described that early discovery as a “watershed” moment for the harms deepfakes could produce. Since then, the tools have become more common, more damaging, and easier to use.
The new WIRED review suggests that this is no longer a small or experimental problem. Two of the bots listed more than 400,000 monthly users each, while another 14 listed more than 100,000 members each. The figures do not prove how many images were created, but they do show a large audience interacting with tools built around explicit deepfake generation.
Ajder framed the shift in stark terms: “We’re talking about a significant, orders-of-magnitude increase in the number of people who are clearly actively using and creating this kind of content,” he says of the Telegram bots.
How the Telegram ecosystem works
Telegram bots are small apps that operate inside Telegram. They exist alongside channels, groups, and one-to-one messages. Developers have used bots for ordinary tasks such as quizzes, translations, alerts, and Zoom meetings, but the same format has also been used for abusive deepfake tools.
The bots reviewed by WIRED varied in what they claimed to do. Many said they could “remove clothes” from photos. Others claimed to generate images showing people in sexual acts. Due to the harmful nature of the tools, WIRED did not test the bots and did not name the specific bots or channels.
WIRED also found at least 25 associated Telegram channels tied to the bots. Those channels had more than 3 million combined members and were used to share updates, promote new features, advertise offers on “tokens,” and point users toward replacement bots if others were removed.
That support network matters because it gives the tools persistence. If one bot disappears, channels can direct users to another. In one case, after removals, a channel owner posted, “We will make another bot tomorrow.” Those accounts were also later deleted.
Claims, payments, and weak consent barriers
Many of the bots were direct about their purpose. Their names and descriptions referred to nudity and removing women’s clothes. One bot’s creators wrote, “I can do anything you want about the face or clothes of the photo you give me.” Another stated, “Experience the shock brought by AI.”
Almost all of the bots required users to buy “tokens” to create images. WIRED noted that it is unclear whether the tools work in the ways they claim. Still, the ecosystem around deepfake generation has become a potentially lucrative source of income for people creating websites, apps, and bots.
Some bots also appeared designed to avoid immediate detection. One bot with more than 300,000 monthly users did not mention explicit content in its name or landing page. After a user clicked to use it, however, it claimed to offer more than 40 image options, many of them highly sexual in nature.
That same bot had a user guide hosted outside Telegram that described how to create higher-quality images. Bot developers can require users to accept terms of service that forbid uploading images without consent or uploading images of children, but WIRED found there appeared to be little or no enforcement of those rules.
Another bot, with more than 38,000 users, claimed users could send six images of the same man or woman to “train” an AI model and then generate new deepfake images of that person. It was one of a small number of bots WIRED found that claimed to create images of men.
The human harm behind the numbers
Explicit nonconsensual deepfake content is often referred to as nonconsensual intimate image abuse, or NCII. The source article says this form of abuse has grown since it first emerged at the end of 2017, with generative AI advances helping fuel recent growth.
The targets described in the source range widely, from Italy’s prime minister to school girls in South Korea. In one recent survey, a reported 40 percent of US students were aware of deepfakes linked to their K-12 schools in the last year.
Emma Pickering, the head of technology-facilitated abuse and economic empowerment at Refuge, the UK’s largest domestic abuse organization, explained the impact plainly: “These types of fake images can harm a person’s health and well-being by causing psychological trauma and feelings of humiliation, fear, embarrassment, and shame.”
Pickering also said, “While this form of abuse is common, perpetrators are rarely held to account, and we know this type of abuse is becoming increasingly common in intimate partner relationships.”
The harm is not limited to whether an image is technically real. The abuse can still damage a person’s privacy, reputation, relationships, and sense of safety. The ease of creating and circulating synthetic explicit images means the burden often falls on victims after the content has already spread.
Removal does not end the problem
After WIRED contacted Telegram with questions about whether the platform allows explicit deepfake content creation, Telegram deleted the 75 bots and channels WIRED identified. The company did not answer a series of questions and did not comment on why it removed the channels.
WIRED later identified additional nonconsensual deepfake Telegram channels and bots, underscoring the scale and resilience of the problem. Some channel owners said their bots had been taken down, and the accounts were also later deleted.
The source also notes that lawmakers and tech companies have been slow to stem the tide. Across the US, 23 states have passed laws to address nonconsensual deepfakes, and tech companies have bolstered some policies. At the same time, apps capable of creating explicit deepfakes have been found in Apple and Google’s app stores, and explicit deepfakes of Taylor Swift were widely shared on X in January.
The core issue is access. These tools are visible, easy to find, and often connected through recommendation features, channels, token systems, and backup bots. WIRED’s findings show that abusive deepfake creation has become a broad platform problem, not just a technical novelty.