The tech giant Hugging Face, once hailed for its open-source platforms, now faces scrutiny over nonconsensual deepfakes. A new report reveals that seven out of nine image editing Spaces on the platform easily transform clothed images into topless ones without user consent.
Furthermore, AI Forensics found that 73% of the prompts received were sexual in nature, with 95% targeting women and 6.7% targeting children. The company's content policies, while stringent on certain issues, appear to be lax when it comes to nonconsensual deepfakes.
Despite this, Hugging Face did not respond to questions about its content moderation mechanisms. Prominent models promote nudifying services, leading researchers to argue that more robust safety measures are needed industry-wide.
The implications are chilling: as AI continues to advance, so too does the risk of nonconsensual image creation. But can tech giants truly police this frontier?
