<p>Sexual fantasies are intimate confessions which most don't want getting out. However, when data from a AI girlfriend site — Muah.ai — was breached, this very dreaded scenario came to pass. </p><p>As per a <em>404 Media</em> <a href="https://www.404media.co/hacked-ai-girlfriend-data-shows-prompts-describing-child-sexual-abuse-2/" rel="nofollow">report</a>, a hacker stole a massive database of users' interaction with their sexual partner chatbots. </p><p>A social media handle has <a href="https://x.com/Laughing_Mantis/status/1844433047980277829" rel="nofollow">claimed</a> there are now extortion attempts based off the information obtained during the breach. </p><p>Muah.ai describes itself as a platform where people can engage with their AI-powered companions for NSFW (not suitable for workplace) chats, as well as exchanging photos or engaging in voice interactions. </p><p>Given the sensitive nature of the data, the site made assurances that the communications were encrypted and has said it does not sell data to third parties. </p>.AI will transform sex work but not intimacy.<p>However, as per the hacker, it was no trouble finding a vulnerability to gain access to the database of the platform that has been described as a 'handful of open-source projects duct-taped together.' </p><p>The stolen data shows that the chatbot prompts are linked to email addresses, many of which are personal accounts with the users' real names. </p><p>The breach also showed that there were horrific explicit references to children and child sexual abuse. </p><p>The administrator of Muah.ai said the hack came to their notice a week back and claims it was sponsored by competitors in the 'uncensored AI industry'. <em>Fortune</em> <a href="https://fortune.com/longform/meta-openai-uncensored-ai-companions-child-pornography/" rel="nofollow">reported</a> that there has been a wave of AI sex companions — many children — in the wake of advancements made by OpenAI and Meta. </p><p>As per the administrator, a moderation team is deployed to suspend and delete all child-related chatbots on the card gallery, but it was reported that when two people posted about an underage AI character on the site's Discord server, a moderator apparently told users not to post the content there, but to directly message each other instead. </p><p>The incident throws light on the new breed of uncensored AI apps in the 'grey zone' of generative AI tech — a black-market of sorts that caters to often taboo role-playing scenarios with 'sex chatbots'. In the fringe wildlands of AI-generated content, the potential for abuse is immense as is the risk of one's fantasies becoming public.</p>
<p>Sexual fantasies are intimate confessions which most don't want getting out. However, when data from a AI girlfriend site — Muah.ai — was breached, this very dreaded scenario came to pass. </p><p>As per a <em>404 Media</em> <a href="https://www.404media.co/hacked-ai-girlfriend-data-shows-prompts-describing-child-sexual-abuse-2/" rel="nofollow">report</a>, a hacker stole a massive database of users' interaction with their sexual partner chatbots. </p><p>A social media handle has <a href="https://x.com/Laughing_Mantis/status/1844433047980277829" rel="nofollow">claimed</a> there are now extortion attempts based off the information obtained during the breach. </p><p>Muah.ai describes itself as a platform where people can engage with their AI-powered companions for NSFW (not suitable for workplace) chats, as well as exchanging photos or engaging in voice interactions. </p><p>Given the sensitive nature of the data, the site made assurances that the communications were encrypted and has said it does not sell data to third parties. </p>.AI will transform sex work but not intimacy.<p>However, as per the hacker, it was no trouble finding a vulnerability to gain access to the database of the platform that has been described as a 'handful of open-source projects duct-taped together.' </p><p>The stolen data shows that the chatbot prompts are linked to email addresses, many of which are personal accounts with the users' real names. </p><p>The breach also showed that there were horrific explicit references to children and child sexual abuse. </p><p>The administrator of Muah.ai said the hack came to their notice a week back and claims it was sponsored by competitors in the 'uncensored AI industry'. <em>Fortune</em> <a href="https://fortune.com/longform/meta-openai-uncensored-ai-companions-child-pornography/" rel="nofollow">reported</a> that there has been a wave of AI sex companions — many children — in the wake of advancements made by OpenAI and Meta. </p><p>As per the administrator, a moderation team is deployed to suspend and delete all child-related chatbots on the card gallery, but it was reported that when two people posted about an underage AI character on the site's Discord server, a moderator apparently told users not to post the content there, but to directly message each other instead. </p><p>The incident throws light on the new breed of uncensored AI apps in the 'grey zone' of generative AI tech — a black-market of sorts that caters to often taboo role-playing scenarios with 'sex chatbots'. In the fringe wildlands of AI-generated content, the potential for abuse is immense as is the risk of one's fantasies becoming public.</p>