~/AI ETHICS/cctv-exposes-black-market-for-ai-undressing-software-and-tutorials

CCTV Exposes Black Market for "AI Undressing" Software and Tutorials

China's state broadcaster CCTV exposed a black market industry chain where "AI undressing" software, tutorials, and prompt documents are sold in online groups to generate non-consensual explicit images from normal photos. A suspect, Zhang, was arrested for making over 100,000 RMB by packaging open-source AI models trained with specific keywords for this purpose. This exposure highlights the growing threat of deepfake technology being weaponized for extortion, defamation, and fraud, prompting calls for stricter enforcement of generative AI regulations. It underscores the ethical and legal challenges of open-source AI models being repurposed for malicious activities. The "undressing" tool does not actually remove clothing but uses inpainting and image-generation algorithms to redraw specific body parts based on prompts. The process is highly automated, allowing users to batch-generate fake explicit images within seconds to minutes.

## BACKGROUND

AI image inpainting is a technology that allows users to select a specific region of an image and use generative AI to modify, replace, or fill in that area based on text prompts. When applied maliciously to generate non-consensual explicit images, this technology is often referred to as deepfake pornography, which has raised global concerns regarding privacy, consent, and digital safety.

## REFERENCES

## KEYWORDS

#AI Ethics#Deepfakes#AI Safety#Regulation

$ subscribe --daily

CCTV Exposes Black Market for "AI Undressing" Software and Tutorials | Daily News