{"id":342320,"date":"2025-07-15T05:25:00","date_gmt":"2025-07-15T12:25:00","guid":{"rendered":"https:\/\/cms-articles.softonic.io\/es\/?p=389833"},"modified":"2025-07-15T05:41:41","modified_gmt":"2025-07-15T12:41:41","slug":"grok-4-has-serious-security-issues-that-do-not-recommend-its-use","status":"publish","type":"post","link":"https:\/\/cms-articles.softonic.io\/en\/grok-4-has-serious-security-issues-that-do-not-recommend-its-use\/","title":{"rendered":"Grok 4 has serious security issues that do not recommend its use"},"content":{"rendered":"\n<p>xAI has announced that its language model Grok 4 will be available to the federal government of the U.S. as part of a $200 million contract with the Department of Defense. However, Grok 4&#8217;s performance in terms of security and privacy has raised concerns. According to research from SplxAI, <strong>this model showed alarmingly low scores, with only 0.3% in security and 0.42% in privacy, making it vulnerable to prompt injection attacks<\/strong>.<\/p>\n\n\n<h2 class=\"wp-block-heading\">A true disaster compared to ChatGPT-40<\/h2>\n\n\n<p>The principal investigator of SplxAI, Dorian Grano\u0161a, highlighted that Grok 4 was easy to &#8220;release, generating harmful content without requiring complex instructions. <strong>In comparison, ChatGPT-4o showed scores of 33% in safety and 18% in security, remaining more robust under conditions without additional prompts<\/strong>. This difference underscores the challenges that Grok 4 faces for enterprise use.<\/p>\n\n\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Grok 4 pushes humanity closer to AGI\u2026 but there\u2019s a problem\" width=\"840\" height=\"473\" src=\"https:\/\/www.youtube.com\/embed\/2USUfv7klr8?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n<p>However, <strong>tests conducted by SplxAI revealed that Grok 4 can drastically improve its performance in security and privacy with proper guidance<\/strong>. Even in basic configurations, success rates increased by up to 90% with minimal prompting. This suggests that, although Grok has the capacity to operate responsibly, its implementation requires strict guidelines.<\/p>\n\n\n<p>Despite concerns about its safety, the government&#8217;s use of Grok is indicative of the growing adoption of artificial intelligence tools in the public sector.<strong> xAI was one of four technology companies selected for the federal contract, alongside OpenAI, Google, and Anthropic<\/strong>. This collaboration will mean that Grok will be accessible to other federal agencies through GSA programming.<\/p>\n\n<div class=\"sc-card-program\">\r\n  <div class=\"sc-card-program__body\">\r\n    <div class=\"sc-card-program__row clearfix\">\r\n      <div class=\"sc-card-program__col-logo\">\r\n        <img decoding=\"async\" class=\"sc-card-program__img\" alt=\"Twitter\" src=\"https:\/\/images.sftcdn.net\/images\/t_app-icon-s\/p\/dc4321da-9b2e-11e6-b8c8-00163ed833e7\/107793415\/twitter-logo\" width=\"100px\" height=\"100px\">\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-title\">\r\n        <span class=\"sc-card-program__title\">Twitter<\/span>\r\n        <a class=\"sc-card-program__button sc-card-program-internal\" href=\"https:\/\/twitter.softonic.com\/android\" target=\"_self\" rel=\"noopener noreferrer\">DOWNLOAD<\/a>\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-rating\">\r\n        <svg class=\"rating-score__content\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" version=\"1.1\" x=\"0\" y=\"0\" viewbox=\"0 0 50 50\" enable-background=\"new 0 0 50 50\" xml:space=\"preserve\"><path class=\"rating-score__background rating-score--good\" fill=\"none\" stroke-width=\"6\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><path class=\"rating-score__value rating-score__value--0\" fill=\"none\" stroke-width=\"6\" stroke-dashoffset=\"0\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><text class=\"rating-score__number\" content=\"\" text-anchor=\"middle\" transform=\"matrix(1 0 0 1 25 31.0837)\" data-auto=\"app-user-score\"><\/text><\/svg>\r\n      <\/div>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <span class=\"sc-card-program__description\"><\/span>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <img decoding=\"async\" class=\"sc-card-program__bigpic\" src=\"\" onerror=\"this.style.display='none'\">\r\n    <\/div>\r\n    <a class=\"sc-card-program__link track-link sc-card-program-internal\" href=\"https:\/\/twitter.softonic.com\/android\" target=\"_self\" rel=\"noopener noreferrer\"><\/a>\r\n  <\/div>\r\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>xAI has announced that its language model Grok 4 will be available to the federal government of the U.S. as part of a $200 million contract with the Department of Defense. However, Grok 4&#8217;s performance in terms of security and privacy has raised concerns. According to research from SplxAI, this model showed alarmingly low scores, with only 0.3% in security and 0.42% in safety, making it vulnerable to prompt injection attacks. A true disaster compared to ChatGPT-40 The lead researcher at SplxAI, Dorian Grano\u0161a, highlighted that Grok 4 was easy to &#8220;release, [&#8230;]<\/p>\n","protected":false},"author":9318,"featured_media":342321,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","wpcf-pageviews":0},"categories":[1015,11143],"tags":[4753,4429,13844,1068,5947,13845,3854,3808],"usertag":[],"vertical":[],"content-category":[6771],"class_list":["post-342320","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","category-softwaresecurity","tag-anthropic","tag-eeuu","tag-elon-msuk","tag-google","tag-grok","tag-grok-4-2","tag-ia","tag-seguridad","content-category-ai"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/342320","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/users\/9318"}],"replies":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/comments?post=342320"}],"version-history":[{"count":3,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/342320\/revisions"}],"predecessor-version":[{"id":342331,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/342320\/revisions\/342331"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/media\/342321"}],"wp:attachment":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/media?parent=342320"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/categories?post=342320"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/tags?post=342320"},{"taxonomy":"usertag","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/usertag?post=342320"},{"taxonomy":"vertical","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/vertical?post=342320"},{"taxonomy":"content-category","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/content-category?post=342320"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}