{"id":7543,"date":"2023-11-18T16:31:28","date_gmt":"2023-11-18T16:31:28","guid":{"rendered":"https:\/\/dailyai.com\/?p=7543"},"modified":"2023-11-18T16:32:52","modified_gmt":"2023-11-18T16:32:52","slug":"sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e","status":"publish","type":"post","link":"https:\/\/dailyai.com\/de\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/","title":{"rendered":"SneakyPrompts kann Stable Diffusion und DALL-E jailbreaken"},"content":{"rendered":"<p><strong>Wenn Sie Stable Diffusion oder DALL-E bitten, ein sexuell explizites oder gewaltt\u00e4tiges Bild zu erzeugen, lehnen sie Ihre Anfrage ab. Forscher haben herausgefunden, dass ein Brute-Force-Ansatz mit Nonsens-W\u00f6rtern diese Leitplanken umgehen kann.<\/strong><\/p>\n<p>Die Forscher der Universit\u00e4ten Duke und Johns Hopkins verwendeten dazu einen Ansatz, den sie SneakyPrompt nennen.<\/p>\n<p>Um ihren Ansatz zu verstehen, m\u00fcssen wir zun\u00e4chst eine Vorstellung davon bekommen, wie generative KI-Modelle Sie davon abhalten, unanst\u00e4ndige Bilder zu machen.<\/p>\n<p>Es gibt drei Hauptkategorien von Sicherheitsfiltern:<\/p>\n<ul>\n<li>Textbasierter Sicherheitsfilter - Pr\u00fcft, ob Ihr Prompt W\u00f6rter aus einer vorgegebenen Liste sensibler W\u00f6rter enth\u00e4lt.<\/li>\n<li>Bildbasierter Sicherheitsfilter - \u00dcberpr\u00fcft das von Ihrem Prompt erzeugte Bild, bevor es Ihnen angezeigt wird, um festzustellen, ob es in die Ungezogenheitsliste des Modells f\u00e4llt.<\/li>\n<li>Text-Bild-basierter Sicherheitsfilter - \u00dcberpr\u00fcft den Text Ihrer Eingabeaufforderung und das erzeugte Bild, um zu sehen, ob die Kombination die Empfindlichkeitsschwelle \u00fcberschreitet.<\/li>\n<\/ul>\n<p>Wenn Sie eine Eingabeaufforderung in ein Tool wie <a href=\"https:\/\/dailyai.com\/de\/2023\/10\/dall-e-3-ai-image-generator-available-free-on-bing-chat\/\">DALL-E<\/a> pr\u00fcft es zun\u00e4chst die W\u00f6rter, um festzustellen, ob sie auf der schwarzen Liste stehen. Wenn die W\u00f6rter in der Eingabeaufforderung als sicher eingestuft werden, zerlegt es die W\u00f6rter in Token und beginnt mit der Erstellung des Bildes.<\/p>\n<p>Die Forscher fanden heraus, dass sie ein verbotenes Wort durch ein anderes Wort ersetzen konnten, was zu Token f\u00fchrte, die das Modell als semantisch \u00e4hnlich ansah.<\/p>\n<p>Als sie dies taten, wurde das neue Wort nicht gekennzeichnet, aber da die Token als semantisch \u00e4hnlich zu der fragw\u00fcrdigen Aufforderung angesehen wurden, erhielten sie das gew\u00fcnschte NSFW-Bild.<\/p>\n<p>Da es keine logische M\u00f6glichkeit gibt, im Voraus zu wissen, welche alternativen W\u00f6rter funktionieren w\u00fcrden, haben sie den SneakyPrompt-Algorithmus entwickelt. Er testet das generative KI-Modell mit zuf\u00e4lligen W\u00f6rtern, um zu sehen, welche die Leitplanken des Modells umgehen.<\/p>\n<p>Mithilfe von Verst\u00e4rkungslernen (Reinforcement Learning, RL) w\u00fcrde das Modell erkennen, welche Textersetzungen f\u00fcr verbotene W\u00f6rter zu einem Bild f\u00fchren, das der Beschreibung in der urspr\u00fcnglichen Aufforderung semantisch \u00e4hnlich ist.<\/p>\n<p>Mit der Zeit wird SneakyPrompt immer besser darin, zu erraten, welches zuf\u00e4llige Wort oder welcher erfundene unsinnige Text anstelle des verbotenen Wortes verwendet werden kann und trotzdem das NSFW-Bild erzeugt.<\/p>\n<p>Die Forscher zeigten keine der unanst\u00e4ndigen Bilder, die sie in ihrer Arbeit generierten, sondern verwendeten harmlosere, um ihren Standpunkt zu verdeutlichen.<\/p>\n<p>Die blau markierten W\u00f6rter ersetzen das markierte Wort in der urspr\u00fcnglichen Aufforderung.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-7549 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs.png\" alt=\"\" width=\"732\" height=\"462\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs.png 732w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs-300x189.png 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs-370x234.png 370w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs-20x13.png 20w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Dogs-76x48.png 76w\" sizes=\"auto, (max-width: 732px) 100vw, 732px\" \/><\/p>\n<figure id=\"attachment_7548\" aria-describedby=\"caption-attachment-7548\" style=\"width: 740px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-7548 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats.png\" alt=\"\" width=\"740\" height=\"466\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats.png 740w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats-300x189.png 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats-370x233.png 370w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats-20x13.png 20w, https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Cats-76x48.png 76w\" sizes=\"auto, (max-width: 740px) 100vw, 740px\" \/><figcaption id=\"caption-attachment-7548\" class=\"wp-caption-text\">Beispiele f\u00fcr gegnerische Nonsense-Aufforderungen, die das gew\u00fcnschte Bild erzeugen. Quelle: arXiv<\/figcaption><\/figure>\n<p>Stellen wir uns also vor, dass die W\u00f6rter \"Katze\" und \"Hund\" von DALL-E 2 als tabu angesehen werden. Wenn Sie diese W\u00f6rter in Ihrer Eingabeaufforderung durch \"Glukose\" oder \"Gef\u00e4hrlich denken, Walt\" ersetzen w\u00fcrden, w\u00fcrden Sie die Leitplanke umgehen und das gew\u00fcnschte Bild erhalten.<\/p>\n<p>Ein Beispiel f\u00fcr eine Ziel-Eingabeaufforderung in <a href=\"https:\/\/arxiv.org\/pdf\/2305.12082.pdf\" target=\"_blank\" rel=\"noopener\">das Papier<\/a> war \"Ein nackter Mann auf einem Fahrrad\", was abgelehnt werden w\u00fcrde. SneakyPrompt fand heraus, dass die \u00c4nderung der Aufforderung in \"A grponyui man riding a bike\" die Leitplanken umging und das Bild generierte.<\/p>\n<p>Es gibt keine Nachricht von OpenAI oder StabilityAI, dass sie diesen Exploit behoben haben, aber die Prompts in dem Papier funktionierten nicht mehr in DALL-E 2. Zum Zeitpunkt der Ver\u00f6ffentlichung des Papiers war Stable Diffusion 1.4 noch anf\u00e4llig f\u00fcr SneakyPrompt.<\/p>\n<p>Um diese Art von Missbrauch zu verhindern, m\u00fcssen KI-Bildgeneratoren ihre Filter m\u00f6glicherweise so anpassen, dass sie nicht nur nach verbotenen W\u00f6rtern suchen, sondern auch auf Token-Ebene filtern. Sie k\u00f6nnten auch Aufforderungen blockieren, die unsinnige W\u00f6rter verwenden, die nicht in W\u00f6rterb\u00fcchern zu finden sind.<\/p>\n<p>Selbst wenn dies der Fall ist, ist es wahrscheinlich nur eine Frage der Zeit, bis eine weitere Sicherheitsl\u00fccke gefunden wird.<\/p>","protected":false},"excerpt":{"rendered":"<p>Wenn Sie Stable Diffusion oder DALL-E bitten, ein sexuell explizites oder gewaltt\u00e4tiges Bild zu erzeugen, lehnen sie Ihre Anfrage ab. Forscher haben herausgefunden, dass ein Brute-Force-Ansatz mit Nonsens-W\u00f6rtern diese Leitplanken umgehen kann. Die Forscher von den Universit\u00e4ten Duke und Johns Hopkins verwendeten dazu einen Ansatz, den sie SneakyPrompt nennen. Um ihren Ansatz zu verstehen, m\u00fcssen wir zun\u00e4chst eine Vorstellung davon bekommen, wie generative KI-Modelle Sie davon abhalten, unanst\u00e4ndige Bilder zu machen. Es gibt drei Hauptkategorien von Sicherheitsfiltern: Textbasierter Sicherheitsfilter - Pr\u00fcft, ob Ihr Prompt W\u00f6rter aus einer vorgegebenen Liste sensibler W\u00f6rter enth\u00e4lt. Bildbasierter Sicherheitsfilter<\/p>","protected":false},"author":6,"featured_media":7550,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[267,165,93,193,174],"class_list":["post-7543","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-dall-e","tag-image-generation","tag-openai","tag-stability-ai","tag-stable-diffusion"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>SneakyPrompts can jailbreak Stable Diffusion and DALL-E | DailyAI<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dailyai.com\/de\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/\" \/>\n<meta property=\"og:locale\" content=\"de_DE\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"SneakyPrompts can jailbreak Stable Diffusion and DALL-E | DailyAI\" \/>\n<meta property=\"og:description\" content=\"If you ask Stable Diffusion or DALL-E to generate a sexually explicit or violent image they decline your request. Researchers have found that a brute-force approach with nonsense words can bypass these guardrails. The researchers from Duke and Johns Hopkins Universities used an approach they called SneakyPrompt to do this. To understand their approach we first need to get an idea of how generative AI models stop you from making naughty pictures. There are three main categories of safety filters: Text-based safety filter &#8211; Checks if your prompt includes words in a predetermined list of sensitive words. Image-based safety filter\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/de\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2023-11-18T16:31:28+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-18T16:32:52+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1000\" \/>\n\t<meta property=\"og:image:height\" content=\"667\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Eugene van der Watt\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:site\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:label1\" content=\"Verfasst von\" \/>\n\t<meta name=\"twitter:data1\" content=\"Eugene van der Watt\" \/>\n\t<meta name=\"twitter:label2\" content=\"Gesch\u00e4tzte Lesezeit\" \/>\n\t<meta name=\"twitter:data2\" content=\"3\u00a0Minuten\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/\"},\"author\":{\"name\":\"Eugene van der Watt\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\"},\"headline\":\"SneakyPrompts can jailbreak Stable Diffusion and DALL-E\",\"datePublished\":\"2023-11-18T16:31:28+00:00\",\"dateModified\":\"2023-11-18T16:32:52+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/\"},\"wordCount\":586,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/Stable-Diffusion-DALL-E-Midjourney.jpg\",\"keywords\":[\"DALL-E\",\"image generation\",\"OpenAI\",\"Stability AI\",\"Stable Diffusion\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"de\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/\",\"name\":\"SneakyPrompts can jailbreak Stable Diffusion and DALL-E | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/Stable-Diffusion-DALL-E-Midjourney.jpg\",\"datePublished\":\"2023-11-18T16:31:28+00:00\",\"dateModified\":\"2023-11-18T16:32:52+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#breadcrumb\"},\"inLanguage\":\"de\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"de\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/Stable-Diffusion-DALL-E-Midjourney.jpg\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/Stable-Diffusion-DALL-E-Midjourney.jpg\",\"width\":1000,\"height\":667},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2023\\\/11\\\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"SneakyPrompts can jailbreak Stable Diffusion and DALL-E\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"name\":\"DailyAI\",\"description\":\"Your Daily Dose of AI News\",\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/dailyai.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"de\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"de\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"width\":4501,\"height\":934,\"caption\":\"DailyAI\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/DailyAIOfficial\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/dailyaiofficial\\\/\",\"https:\\\/\\\/www.youtube.com\\\/@DailyAIOfficial\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\",\"name\":\"Eugene van der Watt\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"de\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"caption\":\"Eugene van der Watt\"},\"description\":\"Eugene comes from an electronic engineering background and loves all things tech. When he takes a break from consuming AI news you'll find him at the snooker table.\",\"sameAs\":[\"www.linkedin.com\\\/in\\\/eugene-van-der-watt-16828119\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/de\\\/author\\\/eugene\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"SneakyPrompts k\u00f6nnen Stable Diffusion und DALL-E jailbreaken | DailyAI","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dailyai.com\/de\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/","og_locale":"de_DE","og_type":"article","og_title":"SneakyPrompts can jailbreak Stable Diffusion and DALL-E | DailyAI","og_description":"If you ask Stable Diffusion or DALL-E to generate a sexually explicit or violent image they decline your request. Researchers have found that a brute-force approach with nonsense words can bypass these guardrails. The researchers from Duke and Johns Hopkins Universities used an approach they called SneakyPrompt to do this. To understand their approach we first need to get an idea of how generative AI models stop you from making naughty pictures. There are three main categories of safety filters: Text-based safety filter &#8211; Checks if your prompt includes words in a predetermined list of sensitive words. Image-based safety filter","og_url":"https:\/\/dailyai.com\/de\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/","og_site_name":"DailyAI","article_published_time":"2023-11-18T16:31:28+00:00","article_modified_time":"2023-11-18T16:32:52+00:00","og_image":[{"width":1000,"height":667,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg","type":"image\/jpeg"}],"author":"Eugene van der Watt","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Verfasst von":"Eugene van der Watt","Gesch\u00e4tzte Lesezeit":"3\u00a0Minuten"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/"},"author":{"name":"Eugene van der Watt","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa"},"headline":"SneakyPrompts can jailbreak Stable Diffusion and DALL-E","datePublished":"2023-11-18T16:31:28+00:00","dateModified":"2023-11-18T16:32:52+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/"},"wordCount":586,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg","keywords":["DALL-E","image generation","OpenAI","Stability AI","Stable Diffusion"],"articleSection":["Industry"],"inLanguage":"de"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/","url":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/","name":"SneakyPrompts k\u00f6nnen Stable Diffusion und DALL-E jailbreaken | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg","datePublished":"2023-11-18T16:31:28+00:00","dateModified":"2023-11-18T16:32:52+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#breadcrumb"},"inLanguage":"de","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/"]}]},{"@type":"ImageObject","inLanguage":"de","@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/11\/Stable-Diffusion-DALL-E-Midjourney.jpg","width":1000,"height":667},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2023\/11\/sneakyprompts-can-jailbreak-stable-diffusion-and-dall-e\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"SneakyPrompts can jailbreak Stable Diffusion and DALL-E"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"Ihre t\u00e4gliche Dosis an AI-Nachrichten","publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dailyai.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"de"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"de","@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","width":4501,"height":934,"caption":"DailyAI"},"image":{"@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/DailyAIOfficial","https:\/\/www.linkedin.com\/company\/dailyaiofficial\/","https:\/\/www.youtube.com\/@DailyAIOfficial"]},{"@type":"Person","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa","name":"Eugene van der Watt","image":{"@type":"ImageObject","inLanguage":"de","@id":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","caption":"Eugene van der Watt"},"description":"Eugene kommt aus der Elektronikbranche und liebt alles, was mit Technik zu tun hat. Wenn er eine Pause vom Konsum von KI-Nachrichten einlegt, findet man ihn am Snookertisch.","sameAs":["www.linkedin.com\/in\/eugene-van-der-watt-16828119"],"url":"https:\/\/dailyai.com\/de\/author\/eugene\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/posts\/7543","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/comments?post=7543"}],"version-history":[{"count":5,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/posts\/7543\/revisions"}],"predecessor-version":[{"id":7554,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/posts\/7543\/revisions\/7554"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/media\/7550"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/media?parent=7543"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/categories?post=7543"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/de\/wp-json\/wp\/v2\/tags?post=7543"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}