Editorial illustration for Study finds ‘bot’ term used 16,232 times in 2.8M Telegram messages
Bots Dominate 2.8M Telegram Messages in EU Study
Study finds ‘bot’ term used 16,232 times in 2.8M Telegram messages
Sixteen thousand two hundred and thirty two. That's the exact number of times users typed "bot" into nearly 2.8 million messages in Italian and Spanish Telegram groups. It's not chatter. It's a transaction log for an industrial abuse factory.
The word "bot" appears 16,232 times across the nearly 2.8 million messages analyzed from 16 Italian and Spanish Telegram groups and channels. Among the disguised redirect links users shared within these groups, 49.71 percent led to AI girlfriend generators and 19.14 percent pointed to nudifying bots, the study found. The researchers also documented how users share prompts for commercial chatbots like Grok or Gemini to manipulate images.
Under the hashtag #PornoTok, users deliberately create synthetic content of female TikTok influencers, including fake audio clips featuring their voices. According to the report, AI has lowered the technical barrier so far that the number of potential victims is growing dramatically.
The data points form a clear pipeline. Almost half of all disguised links lead to AI girlfriend services. One in five goes straight to a tool for stripping clothes from images.
This is the boring, automated back-end of harassment. Users trade prompts to weaponize polished corporate chatbots from xAI and Google. Under #PornoTok, they run production lines for fake porn of female influencers, complete with fabricated audio.
Skill is no longer a prerequisite. The only remaining barrier is intent. Telegram provides the venue, AI provides the tools, and the result is a scalable system for harming people.
The study counts the word "bot" 16,232 times. Each one represents a choice to use a machine to hurt someone.
Related Reading
Research & Benchmarks Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks Databricks finds multi-step agents beat single-turn RAG by 21% to 38% on STaRK
Research & Benchmarks Nvidia's DLSS 4.5 beta adds 6x Multi Frame Generation for RTX 50 GPUs
From Archives Google AI Overviews answers 91% of test questions correctly after Gemini 3 update
From Archives MaxToki AI boosts context to 16,384 tokens with RoPE scaling
Common Questions Answered
How many times was the term 'bot' mentioned in the Telegram messages analyzed?
The study found the term 'bot' appeared 16,232 times across nearly 2.8 million messages from 16 Italian and Spanish Telegram groups and channels. This high frequency suggests a significant and organized presence of automated accounts and bot-related discussions.
What types of links were most commonly shared in these Telegram groups?
According to the research, 49.71 percent of disguised redirect links led to AI girlfriend generators, while 19.14 percent pointed to nudifying bots. These links represent a concerning trend of potentially exploitative and non-consensual digital content generation.
What economic incentives exist for creating nudifying bots?
The study revealed that archives of non-consensual intimate images can sell for 20 to 50 euros, with affiliate marketers claiming up to a 40 percent commission. This financial motivation appears to drive the creation and distribution of synthetic nude image generation technologies.
Further Reading
- Papers with Code Benchmarks — Papers with Code
- Chatbot Arena Leaderboard — LMSYS
