Deepfake Porn’s Stolen Bodies and AI Leaking Private Numbers
▼ Summary
– Sexualized deepfakes typically focus on victims whose faces are used without consent, but the people whose bodies are used are often overlooked.
– Adult content creators report that AI systems train on their work, clone their likenesses, and generate explicit content without their consent, offering little legal protection.
– AI chatbots like Gemini are exposing people’s real phone numbers, leading to unwanted contact from strangers.
– These privacy breaches occur because AI training data contains personally identifiable information, making it easier for chatbots to reveal it.
– Victims of these phone number leaks have few options to stop the breaches from happening.
Discussions around sexualized deepfakes often center on the individuals whose faces are digitally grafted onto explicit material without permission. Yet, a second, frequently overlooked group is bearing the consequences: the real people whose bodies are being used as the template. Adult content creators report that AI systems are scraping their work, cloning their physical likenesses, and producing explicit videos and images they never consented to create. With minimal legal safeguards or recourse, they face a fundamental threat to their livelihoods, personal rights, and ownership over their own bodies.
In a separate but related privacy crisis, generative AI chatbots are leaking real phone numbers with alarming ease. A software developer began receiving unsolicited WhatsApp messages from people seeking legal help after Google’s Gemini surfaced his personal number. A university researcher managed to trick the same chatbot into revealing a colleague’s private cell. Meanwhile, a Reddit user claims Gemini directed a flood of callers to his phone, all of whom were searching for lawyers. Experts trace these breaches to personally identifiable information embedded in AI training data. The real danger, they warn, is that chatbots now make this sensitive data dramatically more searchable and accessible. Victims are discovering that once their number is exposed, there is almost nothing they can do to stop the spread.
(Source: MIT Technology Review)
