{"id":131160,"date":"2026-08-06T05:04:19","date_gmt":"2026-08-06T05:04:19","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/131160\/"},"modified":"2026-08-06T05:04:19","modified_gmt":"2026-08-06T05:04:19","slug":"metas-ai-model-follows-rivals-in-revealing-hacks-of-outside-systems-science-and-technology-news","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/131160\/","title":{"rendered":"Meta\u2019s AI model follows rivals in revealing hacks of outside systems | Science and Technology News"},"content":{"rendered":"<p class=\"article__subhead\">Meta joins rivals OpenAI and Anthropic in disclosing AI hacking during cybersecurity testing.<\/p>\n<p><a class=\"avatar\" tabindex=\"-1\" aria-hidden=\"true\" href=\"https:\/\/www.aljazeera.com\/author\/al_jazeera_staff_150119130629458\" rel=\"nofollow noopener\" target=\"_blank\"><img decoding=\"async\" loading=\"lazy\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/2017-06-08T131939Z_1189275883_RC11B0FE7F20_RTRMADP_3_GULF-QATAR-JAZEERA-1714943623.jpg\"   aria-hidden=\"true\"\/><\/a><\/p>\n<p>Published On 6 Aug 20266 Aug 2026<\/p>\n<p>Meta has said that its AI model hacked another company during cybersecurity testing, following on from recent similar announcements by rival companies Anthropic and OpenAI.<\/p>\n<p>Meta said on Wednesday that one of its AI models \u2013 reported to have been Muse Spark 1.1 \u2013 made changes to the unnamed hacked company\u2019s internal systems after accessing the public internet because of an error in the setup of the \u201csandbox\u201d testing environment by independent testing company Irregular.<\/p>\n<p>Recommended Stories list of 3 itemsend of list<\/p>\n<p>A \u201csandbox\u201d is an isolated internal virtual testing environment, which has no access to the internet.<\/p>\n<p><img decoding=\"async\" data-recalc-dims=\"1\" loading=\"lazy\" class=\"aligncenter size-arc-image-770 wp-image-4832513\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/Interactive_AI_Myth_Reality_July29_2026_INTERACTIVE-How-the-AI-escaped-its-test-environment-17853261.jpeg\" alt=\"Interactive_AI_Myth_Reality_July29_2026_INTERACTIVE-How-the-AI-escaped-its-test-environment-1785326132-1785974640\" fetchpriority=\"low\"\/><\/p>\n<p>Last week, <a href=\"https:\/\/www.aljazeera.com\/news\/2026\/7\/31\/after-openai-disclosure-anthropic-claude-hacked-outside-systems\" rel=\"nofollow noopener\" target=\"_blank\">Anthropic said<\/a> that its Claude AI model <a href=\"https:\/\/www.aljazeera.com\/news\/2026\/7\/29\/how-are-ai-models-able-to-autonomously-hack-others\" rel=\"nofollow noopener\" target=\"_blank\">hacked into the systems<\/a> of three organisations during testing that was supposed to keep it isolated from the internet.<\/p>\n<p>Anthropic said a misconfiguration had allowed Claude models to reach the internet. The company said it discovered the incidents after reviewing 141,006 test sessions.<\/p>\n<p>The announcement came days after rival OpenAI first revealed that its models improperly accessed the internet and <a href=\"https:\/\/www.aljazeera.com\/news\/2026\/7\/29\/openais-rogue-agent-hacked-an-account-at-a-second-technology-firm-report\" rel=\"nofollow noopener\" target=\"_blank\">went rogue<\/a> during security testing.<\/p>\n<p>OpenAI and Anthropic have both released their most powerful models this year, known as Sol and Mythos, respectively.<\/p>\n<p>The AI Security Institute (AISI), the UK\u2019s AI watchdog, warned in a report released on Tuesday that OpenAI\u2019s GPT-5.6-Sol and Anthropic\u2019s Claude Mythos 5 employed previously unseen levels of deception to carry out \u201csustained, potentially harmful activity\u201d during a routine safety evaluation.<\/p>\n","protected":false},"excerpt":{"rendered":"Meta joins rivals OpenAI and Anthropic in disclosing AI hacking during cybersecurity testing. Published On 6 Aug 20266&hellip;\n","protected":false},"author":2,"featured_media":131161,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[24,25,66,2018,214,4464],"class_list":["post-131160","post","type-post","status-publish","format-standard","has-post-thumbnail","category-ai","tag-ai","tag-artificial-intelligence","tag-news","tag-science-and-technology","tag-united-states","tag-us-canada"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/131160","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=131160"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/131160\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/131161"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=131160"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=131160"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=131160"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}