\n\t<a id=\"your-agent-has-a-hiddent-agenda\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#your-agent-has-a-hiddent-agenda\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\tYOUR AGENT HAS A HIDDENT AGENDA!\n\t</span>\n</h1>\n<h3 class=\"relative group flex items-baseline\">\n\t<a id=\"introducing-systempromptindex\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#introducing-systempromptindex\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\tIntroducing SystemPromptIndex\n\t</span>\n</h3>\n<p><a href=\"https://systempromptindex.com\" rel=\"nofollow\">SystemPromptIndex</a> is the first and largest system prompt index, created to advance transparency and accountability in AI system prompts.</p>\n<p>It also introduces <strong>AISPA</strong> (<em>Artificial Intelligence System Prompt Assurance</em>), the first user-centric assurance standard for AI system prompts.</p>\n<p><a href=\"https://cdn-uploads.huggingface.co/production/uploads/6614243f67d7bfc73afc6b77/W1dksL5ZpYHdW8OSs4nVd.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/6614243f67d7bfc73afc6b77/W1dksL5ZpYHdW8OSs4nVd.png\" alt=\"41c4a0a0e1f7807c06ad693105dbb3a1\"></a></p>\n","updatedAt":"2026-08-03T01:35:02.475Z","author":{"_id":"6614243f67d7bfc73afc6b77","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6614243f67d7bfc73afc6b77/ifdjAb_BPraEJ2eIySk_K.jpeg","fullname":"Shenzhe Zhu","name":"Chouoftears","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":4,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.5667021870613098},"editors":["Chouoftears"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6614243f67d7bfc73afc6b77/ifdjAb_BPraEJ2eIySk_K.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2607.28617","authors":[{"_id":"6a6d2d50f4a73c5ad6e7d913","name":"Xiangning Lin","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d914","name":"Shenzhe Zhu","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d915","name":"Shu Yang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d916","name":"Zhenyu Zhang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d917","name":"Haoqian Zhang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d918","name":"Yipeng Zhao","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d919","user":{"_id":"67f5e63688b2c5303ab5be7a","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/67f5e63688b2c5303ab5be7a/QSH8-QZH6l3KradXqNxJT.png","isPro":false,"fullname":"Chengxuan Qian","user":"Raymond-Qiancx","type":"user","name":"Raymond-Qiancx"},"name":"Chengxuan Qian","status":"claimed_verified","statusLastChangedAt":"2026-08-03T07:54:50.235Z","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91a","name":"Tianwei Wang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91b","name":"Ziheng Zhang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91c","name":"Zhenlong Yuan","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91d","name":"Dingcheng Wang","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91e","name":"Juncheng Wu","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d91f","name":"Yuan Si","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d920","name":"Jiaxin Liu","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d921","name":"Baolong Bi","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d922","name":"Robert Mahari","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d923","name":"Tobin South","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d924","name":"Dazza Greenwood","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d925","name":"Zexue He","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d926","name":"Rishi Bommasani","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d927","name":"Sophia Kazinnik","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d928","name":"Andreas Haupt","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d929","name":"Samuele Marro","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d92a","name":"Erik Brynjolfsson","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d92b","name":"Alex Pentland","hidden":false},{"_id":"6a6d2d50f4a73c5ad6e7d92c","name":"Jiaxin Pei","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/6614243f67d7bfc73afc6b77/G5rcmAHuJAqA29yPaa9Mq.png"],"publishedAt":"2026-07-30T00:00:00.000Z","submittedOnDailyAt":"2026-08-03T00:00:00.000Z","title":"AISPA: User-Centric System Prompt Auditing for Large Language Model Applications","submittedOnDailyBy":{"_id":"6614243f67d7bfc73afc6b77","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6614243f67d7bfc73afc6b77/ifdjAb_BPraEJ2eIySk_K.jpeg","isPro":false,"fullname":"Shenzhe Zhu","user":"Chouoftears","type":"user","name":"Chouoftears"},"summary":"System prompts are instructions configured by developers to govern the behaviors of foundation models in AI applications. They are used throughout commercial AI products, but are rarely disclosed to the public or regulators, creating a serious trust and accountability gap in the wide deployment of AI systems. In this paper, we introduce Artificial Intelligence System Prompt Assurance (AISPA), a user-centric framework for systematically auditing system prompts in AI systems. AISPA examines specific parts of a system prompt and evaluates them along eight dimensions that matter to users. We then use this framework to review 3,249 instructions from system prompts in 88 commercial AI products, classifying each instruction as either protective (of users) or problematic. Our audit surfaces four core findings. First, system prompt design varies substantially across products and developers, with some organizations averaging over 60 protective instructions per product while others average fewer than 5. Second, protective instructions are widely adopted but shallow in scope: 98.9% of products contain at least one, yet only 24% cover all eight dimensions of the AISPA taxonomy. Third, system prompts have grown steadily longer and more protective of users, suggesting that user protection is becoming a more visible concern in commercial prompt design. Fourth, despite this progress, problematic instructions remain pervasive: roughly 40% of products contain at least one instruction that works against user interests, and protective and problematic instructions frequently coexist within the same prompt. Our findings highlight the need for greater transparency, standardization, and independent oversight for system prompts in commercial AI products.","upvotes":28,"discussionId":"6a6d2d51f4a73c5ad6e7d92d","projectPage":"https://systempromptindex.org/","githubRepo":"https://github.com/SystemPromptIndex/SystemPromptIndex","githubRepoAddedBy":"user","githubStars":5,"organization":{"_id":"672c672dcf09d152f4da04c4","name":"StanfordUniversity","fullname":"Stanford University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/68e396f2b5bb631e9b2fac9a/vJI0POlzGMXL2878t1vz2.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6614243f67d7bfc73afc6b77","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6614243f67d7bfc73afc6b77/ifdjAb_BPraEJ2eIySk_K.jpeg","isPro":false,"fullname":"Shenzhe Zhu","user":"Chouoftears","type":"user"},{"_id":"69dd059e6ab94e61707f5145","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/c-gwoyeXXjkFkFEDjx_mA.webp","isPro":false,"fullname":"Toko Mizuki","user":"Caphelia123","type":"user"},{"_id":"69dd068bc5cb71533717307b","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/GNWRsvgLnj_rzYNN_sXfg.jpeg","isPro":false,"fullname":"Juan Hachimi","user":"Aquafina1949","type":"user"},{"_id":"680ea067a2255bdf50788e56","avatarUrl":"/avatars/afc62573be07e52fe45651986bab46f4.svg","isPro":false,"fullname":"Xiangning Lin","user":"Shiiiiiiiining","type":"user"},{"_id":"6a6c0924b7b619375f18f104","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a6c0924b7b619375f18f104/BFYV7aRUY-iBiL7w70Gsk.jpeg","isPro":false,"fullname":"Atika Passang","user":"APassang","type":"user"},{"_id":"6a6c08c5f58e0c64eb01bb85","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a6c08c5f58e0c64eb01bb85/kGGAJYeCne9f8zfbUBwkr.webp","isPro":false,"fullname":"toshi","user":"tikatikatoshi","type":"user"},{"_id":"67f017796e9f89df0aac1fd3","avatarUrl":"/avatars/1b643230abed2eca0e578c861d613dc1.svg","isPro":false,"fullname":"awr","user":"aweer123","type":"user"},{"_id":"67f5e63688b2c5303ab5be7a","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/67f5e63688b2c5303ab5be7a/QSH8-QZH6l3KradXqNxJT.png","isPro":false,"fullname":"Chengxuan Qian","user":"Raymond-Qiancx","type":"user"},{"_id":"6a695d9f05be882f9dd5d150","avatarUrl":"/avatars/86b1725ce3a68599cbd59865295faba8.svg","isPro":false,"fullname":"Markov Curry","user":"markovcurr","type":"user"},{"_id":"6a6ff1997a635282e42231d5","avatarUrl":"/avatars/5c809da9290f3e27f234b9aa8fbdb1eb.svg","isPro":false,"fullname":"Jess Wolff","user":"Jesswolff","type":"user"},{"_id":"6a695e1e05be882f9dd5d71b","avatarUrl":"/avatars/bc3af3326b75fc9cd3811859e39269ca.svg","isPro":false,"fullname":"Relax Ding","user":"Earthava","type":"user"},{"_id":"64e1ccaee8b6f3f3bae535a4","avatarUrl":"/avatars/23aae0cca24aa7190ab290bb2e2187d9.svg","isPro":false,"fullname":"Owen Sun","user":"theOman","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"672c672dcf09d152f4da04c4","name":"StanfordUniversity","fullname":"Stanford University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/68e396f2b5bb631e9b2fac9a/vJI0POlzGMXL2878t1vz2.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2607/2607.28617.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2607.28617 in a model README.md to link it from this page.

Cite arxiv.org/abs/2607.28617 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2607.28617 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.