Report Finds Hugging Face Models Frequently Enable Nonconsensual Deepfakes
July 28, 2026
A report from AI Forensics indicates seven of the top nine image editing models on Hugging Face can be used to create nonconsensual deepfakes of women and children. The findings highlight insufficient guardrails within the open-source model repository.
HOW THIS AFFECTS YOU
●
builderReview your model's safety alignment and content moderation layers to mitigate misuse risks.
●
policyExpect increased pressure for standardized safety evaluations and hosting regulations for open-source repositories.