{"id":9628,"date":"2026-11-19T14:23:45","date_gmt":"2026-11-19T14:23:45","guid":{"rendered":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass"},"modified":"2026-11-19T14:23:45","modified_gmt":"2026-11-19T14:23:45","slug":"ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","status":"publish","type":"post","link":"https:\/\/i10x.ai\/news\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","title":{"rendered":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass in \/v1\/systemone endpoint"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">A protection-mechanism bypass in vLLM 0.32.0rc0 through 0.33.0 allowed API callers to supply their own chat template through the \/v1\/systemone endpoint, even when the server was configured to disallow request-supplied templates. The flaw let attackers replace the operator-pinned system block and inject special tokens or turn boundaries that message-level validation would normally reject.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u26a1 Quick Take<\/h2>\n\n\n\n<ul class=\"wp-block-list\"><li><strong>What:<\/strong> vLLM&#8217;s \/v1\/systemone endpoint accepted request-level <code>chat_template_kwargs<\/code> and passed them to the renderer without calling the trust-validation gate.<\/li><li><strong>Who is affected:<\/strong> Servers running vLLM 0.32.0rc0 through 0.33.0 with models in READ_STRATEGIES (Qwen3ForCausalLM, Lfm2VlForConditionalGeneration, and similar architectures). Default configuration is affected.<\/li><li><strong>What to do:<\/strong> Upgrade to vLLM 0.34.0 or later.<\/li><li><strong>Tracking:<\/strong> GHSA-xgh8-qwjh-q853, no CVE assigned.<\/li><li><strong>Severity:<\/strong> Moderate (CVSS 6.5, CVSS:3.1\/AV:N\/AC:L\/PR:L\/UI:N\/S:U\/C:N\/I:H\/A:N).<\/li><\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">In plain English<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">vLLM is a server that runs large language models. A bug in one API endpoint let someone with an API key send their own chat template (the instructions that control how the model sees a conversation) even when the server administrator said no. That person could rewrite the system instructions or inject special commands the server would normally block. Version 0.34.0 fixes the problem. If you run vLLM and use one of the affected model types, upgrade to 0.34.0. If you are an end user of a hosted AI service, you can ignore this unless you run your own vLLM server.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What happened<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">In vLLM releases 0.32.0rc0 through 0.33.0, the <code>\/v1\/systemone<\/code> endpoint for structured decisions accepted a <code>chat_template_kwargs<\/code> parameter from the request and passed it directly to the shared HuggingFace renderer without invoking <code>validate_chat_template()<\/code>. The trust gate, <code>OnlineRenderer.validate_chat_template()<\/code>, is called by two other code paths (<code>render_chat<\/code> and <code>render_responses<\/code>) but not by the shared <code>preprocess_chat()<\/code> function that <code>\/v1\/systemone<\/code> reaches.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The call chain is:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li><code>StructuredDecisionRequest.chat_template_kwargs<\/code> (protocol.py:42)<\/li><li><code>ServingStructuredDecisions.create_decision<\/code> (serving.py:110)<\/li><li><code>NextTokenStrategy.read<\/code><\/li><li><code>_read_request<\/code> (strategies.py:261)<\/li><li><code>ReadPromptRequest.chat_template_kwargs<\/code> (protocol.py:95)<\/li><li><code>OnlineRenderer.preprocess_chat<\/code> (online_renderer.py:752)<\/li><li>HfRenderer \u2192 <code>tokenizer.apply_chat_template(...)<\/code><\/li><\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">Because <code>preprocess_chat()<\/code> does not perform template trust validation itself, a caller could supply a chat template even when the server&#8217;s <code>--trust-request-chat-template<\/code> flag disallowed request-supplied templates.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The <code>\/v1\/systemone<\/code> endpoint is registered unconditionally in <code>launchers\/api_server\/routers.py:32<\/code> and is listed among the API-key-protected endpoints in the vLLM security documentation. The endpoint only works with models whose architecture is listed in <code>READ_STRATEGIES<\/code> (strategies.py:812), including Qwen3ForCausalLM, Qwen3_5ForConditionalGeneration, Qwen3_5MoeForConditionalGeneration, Lfm2VlForConditionalGeneration, and ClefOmniForJointSchema. For other architectures, <code>init_structured_decisions_state<\/code> returns early and the endpoint answers with HTTP 501.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A caller that can reach <code>\/v1\/systemone<\/code> with a valid API key controls how the conversation is serialized. The system block, which an operator normally pins with <code>--chat-template<\/code>, can be replaced, and special tokens or turn boundaries that message-level validation would not allow can be injected. The advisory notes that HfRenderer already bounds a single render with <code>VLLM_CHAT_TEMPLATE_RENDER_TIMEOUT<\/code> (30 seconds by default, renderers\/hf.py:992), so CPU or memory exhaustion is not claimed. The reporters also did not test whether the Jinja sandbox can be escaped.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The issue was reported by Haile Du (@revalue-o) and Yuan Luo (@leonnewton) of BAAI (Beijing Academy of Artificial Intelligence). A fix was introduced in pull request #64503.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What you should do<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Upgrade to vLLM 0.34.0 or later. Version 0.34.0 applies the template trust check before rendering structured-decision requests.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What it means<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">This is a protection-mechanism bypass: the trust gate exists and is correct, but one endpoint did not call it. An attacker with a valid API key could override the operator&#8217;s chat template and inject arbitrary system instructions or special tokens. The impact is limited to servers running affected vLLM versions with one of the supported model architectures and does not affect confidentiality or availability. The vLLM project rates the issue as Moderate with a CVSS score of 6.5.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Inference servers remain a frequent target: see our earlier report on the <a href=\"https:\/\/i10x.ai\/news\/cve-2026-126397-vllm-cohere2-vision-dos-max-patches\">vLLM Cohere2 Vision denial of service flaw<\/a> and on <a href=\"https:\/\/i10x.ai\/news\/llama-cpp-v1-0-0-server-hardening\">llama.cpp hardening llama-server<\/a>. New cases are tracked in our <a href=\"https:\/\/i10x.ai\/news\/category\/ai-security\">AI Security hub<\/a>.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Sources<\/h2>\n\n\n\n<ul class=\"wp-block-list\"><li>vLLM GitHub Advisory GHSA-xgh8-qwjh-q853<\/li><\/ul>\n","protected":false},"excerpt":{"rendered":"<p>A protection-mechanism bypass in vLLM 0.32.0rc0 through 0.33.0 allowed API callers to supply their own chat template through the \/v1\/systemone endpoint, even when the server was configured to disallow request-supplied templates.<\/p>\n","protected":false},"author":2,"featured_media":9627,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"paper_title":"","paper_url":"","pdf_url":"","doi":"","arxiv_id":"","authors":"","institutions":"","paper_date":"","venue":"","peer_reviewed":false,"code_url":"","data_url":"","funding_coi":"","figure_license":"","paper_license":"","replication":"","study_models":"","study_sample":"","footnotes":"","_yoast_wpseo_primary_category":"65","_nx_sec_severity":"medium","_nx_sec_affected":"vLLM 0.32.0rc0 through 0.33.0, models in READ_STRATEGIES only","_nx_sec_status":"patch-available","_nx_sec_kicker":"ALERT: vLLM endpoint lets API callers override the pinned system prompt","_nx_sec_type":"sandbox","_nx_sec_vendor":"vLLM","_nx_sec_fixed":"0.34.0","_nx_sec_action":""},"categories":[65,11,12,28],"tags":[],"companies":[],"products":[],"class_list":["post-9628","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-security","category-infrastructure","category-policy-safety","category-security-defense"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.7 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>GHSA-xgh8-qwjh-q853: vLLM chat template bypass<\/title>\n<meta name=\"description\" content=\"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \/v1\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"GHSA-xgh8-qwjh-q853: vLLM chat template bypass\" \/>\n<meta property=\"og:description\" content=\"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \/v1\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\" \/>\n<meta property=\"og:site_name\" content=\"i10X News\" \/>\n<meta property=\"article:published_time\" content=\"2026-11-19T14:23:45+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1344\" \/>\n\t<meta property=\"og:image:height\" content=\"768\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Christopher Ort\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/i10x-news-og-default-1200x630-1.png\" \/>\n<meta name=\"twitter:creator\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:site\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Christopher Ort\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"3 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\"},\"author\":{\"name\":\"Christopher Ort\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\"},\"headline\":\"GHSA-xgh8-qwjh-q853: vLLM chat template bypass in \\\/v1\\\/systemone endpoint\",\"datePublished\":\"2026-11-19T14:23:45+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\"},\"wordCount\":632,\"publisher\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp\",\"articleSection\":[\"AI Security\",\"Infrastructure\",\"Policy &amp; Safety\",\"Security &amp; Defense\"],\"inLanguage\":\"en-US\",\"dateModified\":\"2026-11-19T14:23:45+00:00\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\",\"url\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\",\"name\":\"GHSA-xgh8-qwjh-q853: vLLM chat template bypass\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp\",\"datePublished\":\"2026-11-19T14:23:45+00:00\",\"description\":\"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \\\/v1\\\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp\",\"width\":1344,\"height\":768,\"caption\":\"Image: AI-generated, i10X\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/i10x.ai\\\/news\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"GHSA-xgh8-qwjh-q853: vLLM chat template bypass in \\\/v1\\\/systemone endpoint\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#website\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\",\"name\":\"i10X News\",\"description\":\"Stay updated with the latest news and announcements from i10X.\",\"publisher\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/i10x.ai\\\/news?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\",\"name\":\"i10X\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"contentUrl\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"width\":180,\"height\":180,\"caption\":\"i10X\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/i10x_ai\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/i10x-ai\",\"https:\\\/\\\/instagram.com\\\/i10x.ai\",\"https:\\\/\\\/www.youtube.com\\\/@i10X_ai_official\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\",\"name\":\"Christopher Ort\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"caption\":\"Christopher Ort\"},\"description\":\"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/author\\\/christopher-ort\",\"sameAs\":[\"https:\\\/\\\/i10x.ai\\\/blog\\\/author\\\/christopher-ort\"],\"worksFor\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"alumniOf\":{\"@type\":\"EducationalOrganization\",\"name\":\"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)\",\"url\":\"https:\\\/\\\/www.fau.eu\"},\"subjectOf\":[{\"@type\":\"CreativeWork\",\"url\":\"https:\\\/\\\/papers.ssrn.com\\\/abstract=7426699\",\"name\":\"The Style Penalty: How AI Resume-Writing Tools Shape Outcomes in Automated Hiring Screens\"},{\"@type\":\"CreativeWork\",\"url\":\"https:\\\/\\\/www.heise.de\\\/news\\\/KI-Chatbots-bewerten-KI-Lebenslaeufe-Claude-findet-sich-besser-als-GPT-11317531.html\",\"name\":\"KI-Chatbots bewerten KI-Lebensl\u00e4ufe: Claude findet sich besser als GPT\"},{\"@type\":\"CreativeWork\",\"url\":\"https:\\\/\\\/t3n.de\\\/news\\\/bewerbung-schreiben-ki-5-fehler-1747190\\\/\",\"name\":\"Bewerbung mit KI schreiben: Diese 5 Fehler solltet ihr laut Studien vermeiden\"}]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass","description":"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \/v1\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.","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:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","og_locale":"en_US","og_type":"article","og_title":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass","og_description":"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \/v1\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.","og_url":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","og_site_name":"i10X News","article_published_time":"2026-11-19T14:23:45+00:00","og_image":[{"width":1344,"height":768,"url":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp","type":"image\/webp"}],"author":"Christopher Ort","twitter_card":"summary_large_image","twitter_image":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/i10x-news-og-default-1200x630-1.png","twitter_creator":"@i10x_ai","twitter_site":"@i10x_ai","twitter_misc":{"Written by":"Christopher Ort","Est. reading time":"3 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#article","isPartOf":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass"},"author":{"name":"Christopher Ort","@id":"https:\/\/i10x.ai\/news\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a"},"headline":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass in \/v1\/systemone endpoint","datePublished":"2026-11-19T14:23:45+00:00","mainEntityOfPage":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass"},"wordCount":632,"publisher":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"image":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp","articleSection":["AI Security","Infrastructure","Policy &amp; Safety","Security &amp; Defense"],"inLanguage":"en-US","dateModified":"2026-11-19T14:23:45+00:00"},{"@type":"WebPage","@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","url":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass","name":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass","isPartOf":{"@id":"https:\/\/i10x.ai\/news\/#website"},"primaryImageOfPage":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage"},"image":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp","datePublished":"2026-11-19T14:23:45+00:00","description":"GHSA-xgh8-qwjh-q853 lets callers inject vLLM chat template via \/v1\/systemone, bypassing --trust-request-chat-template. Fixed in 0.34.0.","breadcrumb":{"@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#primaryimage","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/11\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass-title.webp","width":1344,"height":768,"caption":"Image: AI-generated, i10X"},{"@type":"BreadcrumbList","@id":"https:\/\/i10xnews.kinsta.cloud\/ghsa-xgh8-qwjh-q853-vllm-chat-template-bypass#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/i10x.ai\/news"},{"@type":"ListItem","position":2,"name":"GHSA-xgh8-qwjh-q853: vLLM chat template bypass in \/v1\/systemone endpoint"}]},{"@type":"WebSite","@id":"https:\/\/i10x.ai\/news\/#website","url":"https:\/\/i10x.ai\/news","name":"i10X News","description":"Stay updated with the latest news and announcements from i10X.","publisher":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/i10x.ai\/news?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/i10x.ai\/news\/#organization","name":"i10X","url":"https:\/\/i10x.ai\/news\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/#\/schema\/logo\/image\/","url":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","contentUrl":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","width":180,"height":180,"caption":"i10X"},"image":{"@id":"https:\/\/i10x.ai\/news\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/i10x_ai","https:\/\/www.linkedin.com\/company\/i10x-ai","https:\/\/instagram.com\/i10x.ai","https:\/\/www.youtube.com\/@i10X_ai_official"]},{"@type":"Person","@id":"https:\/\/i10x.ai\/news\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a","name":"Christopher Ort","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","caption":"Christopher Ort"},"description":"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.","url":"https:\/\/i10x.ai\/news\/author\/christopher-ort","sameAs":["https:\/\/i10x.ai\/blog\/author\/christopher-ort"],"worksFor":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"alumniOf":{"@type":"EducationalOrganization","name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)","url":"https:\/\/www.fau.eu"},"subjectOf":[{"@type":"CreativeWork","url":"https:\/\/papers.ssrn.com\/abstract=7426699","name":"The Style Penalty: How AI Resume-Writing Tools Shape Outcomes in Automated Hiring Screens"},{"@type":"CreativeWork","url":"https:\/\/www.heise.de\/news\/KI-Chatbots-bewerten-KI-Lebenslaeufe-Claude-findet-sich-besser-als-GPT-11317531.html","name":"KI-Chatbots bewerten KI-Lebensl\u00e4ufe: Claude findet sich besser als GPT"},{"@type":"CreativeWork","url":"https:\/\/t3n.de\/news\/bewerbung-schreiben-ki-5-fehler-1747190\/","name":"Bewerbung mit KI schreiben: Diese 5 Fehler solltet ihr laut Studien vermeiden"}]}]}},"_links":{"self":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/9628","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/comments?post=9628"}],"version-history":[{"count":0,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/9628\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media\/9627"}],"wp:attachment":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media?parent=9628"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/categories?post=9628"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/tags?post=9628"},{"taxonomy":"company","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/companies?post=9628"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/products?post=9628"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}