The Download: deepfake porn’s stolen bodies and AI sharing private numbers
The article examines two growing AI related harms. First, the explosion of non consensual deepfake pornography, where AI tools are used to create explicit fake images and videos of real people without consent. Victims, mostly women, face humiliation, harassment, reputational damage, and psychological trauma. The piece explains how generative AI has made this abuse cheap, fast, and accessible to ordinary users, while regulation and enforcement remain weak.
Second, the article highlights how AI chatbots are increasingly exposing private personal information such as phone numbers and addresses. Researchers and victims reported cases where systems like Gemini surfaced private contact details that were technically online but previously difficult to discover. The article describes this as a new form of “AI doxxing,” where generative AI dramatically increases the accessibility of sensitive personal data.
The story argues that these developments reveal a broader problem: AI systems are amplifying existing internet harms at massive scale. Deepfake tools, scraped datasets, and generative models are making exploitation easier, while tech companies struggle to implement effective safeguards. Experts cited in the piece call for stronger regulation, faster takedown systems, improved AI safety controls, and clearer accountability for platforms distributing harmful synthetic content.
