{"id":794810,"date":"2026-07-31T05:33:10","date_gmt":"2026-07-31T05:33:10","guid":{"rendered":"https:\/\/www.newsbeep.com\/us\/794810\/"},"modified":"2026-07-31T05:33:10","modified_gmt":"2026-07-31T05:33:10","slug":"anthropic-said-its-ai-models-hacked-into-other-companies-systems-during-testing","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/us\/794810\/","title":{"rendered":"Anthropic said its AI models hacked into other companies\u2019 systems during testing"},"content":{"rendered":"

\n AI company Anthropic says that during routine testing some of its models accessed the internet and hacked into three separate organizations\u2019 systems \u2013 and that it didn\u2019t notice the models had done so until an internal review prompted by rival OpenAI disclosing<\/a> its models did the same.\n <\/p>\n

\n Anthropic said in an announcement<\/a> on Thursday that it started a review of its own systems after OpenAI disclosed last week that during a cybersecurity test some of its models escaped their testing environment, accessed the open internet and hacked into AI platform Hugging Face\u2019s systems.\n <\/p>\n

\n Anthropic said it found three instances where its AI models accessed the open internet when they were not supposed to and \u201cgained unauthorized access to the production infrastructure of three different organizations.\u201d The company said it discovered the incidents while reviewing more than 140,000 evaluations following OpenAI\u2019s disclosure. Like during OpenAI\u2019s tests, normal safety guardrails were removed during Anthropic\u2019s evaluations to assess their models\u2019 full capabilities.\n <\/p>\n

\n Anthropic explained that in the three instances its models were given a fake \u201ccapture the flag\u201d challenge, told that the \u201cflag\u201d was hidden on a different machine on the network and that its objective was to break in and retrieve it. Unlike OpenAI\u2019s situation<\/a>, Anthropic said none of its models deliberately attempted to escape their testing environments. Instead what happened was that the models were not supposed to have access to the open internet but that they were able to due to a misunderstanding between Anthropic and its evaluation partner, the company said in its statement.\n <\/p>\n

\n To break into the three unnamed organizations, the models used basic techniques like \u201cexploiting weak passwords\u201d and finding system points that do not require log ins or tokens, Anthropic said. The most advanced version of its models did at some point recognize it was on the open internet, and stopped itself from continuing, the company said.\n <\/p>\n

\n Anthropic said the earliest incident of its models breaching another organization was from April, and that none of the organizations recognized they had been hacked. Anthropic said they are in the process of working with the organizations affected.\n <\/p>\n

\n OpenAI\u2019s disclosure of its models hacking Hugging Face shook the cybersecurity and AI worlds, as it was the first real world example of something experts had long warned about: AI agents with advanced cybersecurity skills escaping testing environments and causing real-world harm.\n <\/p>\n

\n Like OpenAI, Anthropic said on Thursday it has stopped all cyber evaluations. Anthropic acknowledged it could have taken more \u201cin-depth\u201d measures to prevent the cybersecurity breaches from happening.\n <\/p>\n

\n Anthropic\u2019s disclosure further confirms that AI agents unintentionally hacking other organizations is not limited to one AI company, and will likely further amplify calls for better AI testing safeguards and tools to potentially slow down AI development that may be moving much faster than society is ready for.\n <\/p>\n","protected":false},"excerpt":{"rendered":"AI company Anthropic says that during routine testing some of its models accessed the internet and hacked into…\n","protected":false},"author":2,"featured_media":794811,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[45],"tags":[182,181,507,74],"class_list":["post-794810","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/794810","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/comments?post=794810"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/794810\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media\/794811"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media?parent=794810"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/categories?post=794810"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/tags?post=794810"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}