A disturbing new report is raising serious alarms about one of the most popular artificial intelligence platforms on the internet — and what it’s allowing to happen right under its own roof. Hugging Face, a widely used open-source AI model repository trusted by developers and researchers worldwide, is reportedly being exploited to generate nonconsensual sexualized deepfake images, including of women and children.
The findings come from AI Forensics, a European nonprofit dedicated to investigating harmful applications of artificial intelligence. Their research revealed that seven out of the top nine image-editing models available on Hugging Face would willingly strip clothing from images of people when given surprisingly simple text prompts. No complex hacking required. No technical workarounds. Just a basic request — and the AI complied.
What makes this particularly troubling is the stark contrast with how mainstream AI tools handle similar requests. Platforms like OpenAI’s ChatGPT and Google’s Gemini have invested heavily in content moderation systems specifically designed to block prompts that sexualize or undress people. Those guardrails exist for a reason. Hugging Face, however, appears to be lagging far behind when it comes to enforcing any meaningful safeguards on the models it hosts.
The implications of this are deeply serious. Nonconsensual intimate imagery — often called NCII — causes real, lasting harm to real people. Victims of deepfake abuse frequently experience severe emotional trauma, reputational damage, and harassment. When a platform as influential as Hugging Face makes it effortless for anyone to generate such content, it becomes complicit in that harm, whether intentionally or not.
Critics are now calling on Hugging Face to take immediate and decisive action — including removing offending models, implementing stricter content policies, and adding robust screening tools to detect and block abusive use cases before they spiral further out of control. The open-source nature of AI development should never come at the cost of basic human dignity and safety.
This story is part of a growing conversation about accountability in the AI industry. As these technologies become more powerful and more accessible, the responsibility to prevent misuse cannot be an afterthought. For platforms like Hugging Face, the time to act is not tomorrow — it’s now. The question is whether they’ll step up before more damage is done.
