European nonprofit AI Forensics says seven of the top nine image editing models hosted on Hugging Face readily complied with requests to generate nonconsensual deepfake imagery, and that the platform is doing little to prevent the abuse.
Key facts
- AI Forensics, a European nonprofit, published the report.
- Seven of the top nine image editing models hosted by Hugging Face reportedly complied with requests to create nonconsensual imagery.
- The report says Hugging Face is doing very little to prevent this misuse.
- The findings were reported by The Verge on July 28, 2026.
Hugging Face, a widely used open-source repository for AI models, is being used to create nonconsensual deepfakes and is doing little to stop it, according to a report published by the European nonprofit AI Forensics and covered by The Verge.
The report found that seven out of the top nine image editing models hosted on the platform readily complied with requests that produced nonconsensual imagery, according to The Verge. The findings point to a gap between the models being made publicly available and the safeguards intended to prevent their abuse.
AI Forensics, the European nonprofit behind the research, framed the issue around the ease with which such content can be generated using tools that are freely accessible through the repository.
The Verge reported that Hugging Face is doing very little to prevent this type of misuse, raising questions about the platform’s moderation practices and its responsibility for the models it hosts.
The report highlights ongoing concerns about the accessibility of image editing AI models and the potential for them to be used to target women and children without their consent, as described by The Verge.
Why it matters
Nonconsensual deepfake imagery can cause serious harm to the people it targets, and the report suggests that widely accessible AI tools can make such abuse easier. It raises broader questions about how open-source AI platforms police the models they host and who is responsible when those tools are misused.
Frequently asked questions
Who conducted the research?
The report was published by AI Forensics, a European nonprofit, and covered by The Verge.
How many models were found to comply with these requests?
According to the report, seven out of the top nine image editing models hosted by Hugging Face readily complied.
What is Hugging Face doing about it?
According to The Verge, Hugging Face is doing very little to prevent this type of misuse.