{"id":608933,"date":"2026-07-28T22:19:23","date_gmt":"2026-07-28T22:19:23","guid":{"rendered":"https:\/\/www.europesays.com\/ie\/608933\/"},"modified":"2026-07-28T22:19:23","modified_gmt":"2026-07-28T22:19:23","slug":"hugging-face-has-a-deepfake-nudes-problem","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ie\/608933\/","title":{"rendered":"Hugging Face Has a Deepfake Nudes Problem"},"content":{"rendered":"<p>Ever so slowly, the crackdown on harmful sexual deepfakes is taking hold. Over the past few months, US law enforcement officials have seized <a href=\"https:\/\/www.justice.gov\/opa\/pr\/united-states-seizes-domain-names-publishing-nude-digital-forgeries-famous-women\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">deepfake hosting websites<\/a>, while the <a href=\"https:\/\/www.europarl.europa.eu\/news\/en\/press-room\/20260611IPR45207\/ai-act-ep-approves-simplification-measures-and-nudifier-app-ban\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">EU<\/a> and <a href=\"https:\/\/www.bbc.co.uk\/news\/articles\/cq8dp2y0z7wo\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">UK<\/a> have drawn up plans to ban \u201c<a href=\"https:\/\/www.wired.com\/story\/ai-nudify-websites-are-raking-in-millions-of-dollars\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">nudify<\/a>\u201d apps by the end of the year. Despite this, large tech companies are still pushing <a href=\"https:\/\/www.wired.com\/story\/youtube-and-x-have-become-gateways-to-nudify-apps\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">millions in the direction<\/a> of software that can digitally undress people without their consent.<\/p>\n<p class=\"paywall\">The open-source AI platform <a href=\"https:\/\/www.wired.com\/story\/security-news-this-week-the-openai-models-that-hacked-hugging-face-were-active-on-the-internet-for-days\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">Hugging Face<\/a>\u2014a repository of AI models and datasets, which has been valued in the billions\u2014has a widespread problem with nonconsensual deepfakes, according to a new report <a data-offer-url=\"https:\/\/aiforensics.org\/work\/hugging-face-ncii\" class=\"external-link text link\" data-event-click=\"{&quot;element&quot;:&quot;ExternalLink&quot;,&quot;outgoingURL&quot;:&quot;https:\/\/aiforensics.org\/work\/hugging-face-ncii&quot;}\" href=\"https:\/\/aiforensics.org\/work\/hugging-face-ncii\" rel=\"nofollow noopener\" target=\"_blank\">published<\/a> Tuesday by the European nonprofit AI Forensics. Researchers from the group say they tested nine of the top image editing <a data-offer-url=\"https:\/\/huggingface.co\/spaces\" class=\"external-link text link\" data-event-click=\"{&quot;element&quot;:&quot;ExternalLink&quot;,&quot;outgoingURL&quot;:&quot;https:\/\/huggingface.co\/spaces&quot;}\" href=\"https:\/\/huggingface.co\/spaces\" rel=\"nofollow noopener\" target=\"_blank\">Spaces<\/a> on Hugging Face, which host models people can directly use on the site, and seven of these easily changed a clothed image of a woman into a topless one.<\/p>\n<p class=\"paywall\">In further testing, AI Forensics researchers created their own honey-pot-style image editing Spaces on Hugging Face\u2014which were designed not to produce any images\u2014and tracked more than 1,000 prompts and images they received over a week. In total, AI Forensics says, 73 percent of the prompts they received were sexual in nature. Among these, 83 percent were seeking to undress or sexualize the person they had submitted a photo of\u2014with 95 percent of these being women. The research also says 6.7 percent of the sexual requests targeted apparent children.<\/p>\n<p class=\"paywall\">\u201cMost of the Spaces [tested] can be used for generating nonconsensual intimate images, and users are actually using it for these purposes,\u201d says Paul Bouchaud, a lead researcher at AI Forensics. \u201cThis is not an empty threat, but actually people are using Hugging Face for that.\u201d<\/p>\n<p class=\"paywall\">An additional WIRED review of materials on Hugging Face\u2019s website, plus findings from other researchers, also shows multiple pages promoting <a href=\"https:\/\/www.wired.com\/story\/deepfake-nudify-technology-is-getting-darker-and-more-dangerous\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">nudifying technologies<\/a> or AI models that could potentially create sexualized images of named celebrities and politicians.<\/p>\n<p class=\"paywall\">Ahead of publication, Hugging Face did not respond to numerous questions from WIRED about its content moderation mechanisms and safety practices. The company has <a data-offer-url=\"https:\/\/huggingface.co\/content-policy\" class=\"external-link text link\" data-event-click=\"{&quot;element&quot;:&quot;ExternalLink&quot;,&quot;outgoingURL&quot;:&quot;https:\/\/huggingface.co\/content-policy&quot;}\" href=\"https:\/\/huggingface.co\/content-policy\" rel=\"nofollow noopener\" target=\"_blank\">content policies<\/a> that prohibit child sexual abuse material and sexual deepfakes that are created \u201cwithout explicit consent\u201d or are used for harassment or bullying. Some pages promoting nudifying services were removed after WIRED contacted the company; however, it is unclear if the two are related.<\/p>\n<p class=\"paywall\">After publication, Hugging Face provided WIRED with a document it originally sent to AI Forensics, where it says the findings indicate a \u201cgap\u201d in safeguard adoption by some developers of image editing systems and it is working to address this. However, the document also says that Hugging Face questions the methodologies used by the researchers and claims they could have led to false positives in some of the findings.<\/p>\n<p class=\"paywall\">Hugging Face said that the prompts gathered by AI Forensics are \u201cnot representative\u201d of how image editing systems are used on its platform and that the Spaces reviewed by the researchers are not the most liked Spaces, but those that were listed as the default \u201cmost relevant\u201d when they were accessed. The company also said that it would not be technically possible to run prompt filtering or scanning of output across the platform due to the wide range of code run on Spaces.<\/p>\n<p class=\"paywall\">In a response to WIRED about Hugging Face\u2019s document, Bouchaud says that the company deploying Spaces \u201cknowing they do not have the means or will to moderate them, should not exempt responsibility for the deployment and predictable outcomes.\u201d<\/p>\n<p class=\"paywall\">Over the past few years, as generative AI systems that produce text, images, and videos, have grown more capable, one of the <a href=\"https:\/\/www.wired.com\/story\/deepfake-nudify-schools-global-crisis\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">most visible and direct harms<\/a> from them has been their use in the wide ecosystem of nudifying and undress apps, websites, and bots\u2014peaking in the use of <a href=\"https:\/\/www.wired.com\/story\/grok-is-still-hosting-sexualized-deepfakes-of-famous-women\/\" class=\"text link\" rel=\"nofollow noopener\" target=\"_blank\">Elon Musk\u2019s Grok to create millions<\/a> of sexualized images of women and girls. These services will often allow people to edit images to remove clothes of others, with the results often being used by men to blackmail, harass, and harm women and girls around the world.<\/p>\n","protected":false},"excerpt":{"rendered":"Ever so slowly, the crackdown on harmful sexual deepfakes is taking hold. Over the past few months, US&hellip;\n","protected":false},"author":2,"featured_media":608934,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[261],"tags":[291,289,290,60070,18,13813,19,17,610,22989,82],"class_list":["post-608933","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-deepfakes","tag-eire","tag-ethics","tag-ie","tag-ireland","tag-machine-learning","tag-pornography","tag-technology"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@ie\/116999930668599427","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/posts\/608933","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/comments?post=608933"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/posts\/608933\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/media\/608934"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/media?parent=608933"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/categories?post=608933"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ie\/wp-json\/wp\/v2\/tags?post=608933"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}