Anthropic is investigating a claim that a small group of people gained access to its Claude Mythos model<\/b> \u2013 the cyber-security tool which the AI firm says is too powerful to release to the public.
\n“We’re investigating a report claiming unauthorized access to Claude Mythos Preview through one of our third-party vendor environments,” the company said in a statement.
\nIt was in response to a Bloomberg report that users in a private forum managed to access the model without the normal permissions.
\nThere is deep unease about Mythos’ capabilities \u2013 though the UK’s top cyber official has said advanced AI tools could be a “net positive” if the technology was secured from misuse.
\nThere is currently no suggestion that malicious actors have managed to get hold of the model, and Anthropic says it does not have evidence its systems are affected.<\/p>\nConcerns Over AI Model Security<\/h3>\n
But the report of access by unauthorised users raises questions about the ability of large AI companies to stop their advanced AI models from getting into the wrong hands.
\nThis was “most likely through misuse of access rather than a classic hack,” according to Raluca Saceanu, chief executive of cyber-security company Smarttech247.
\nAnthropic has released the Mythos model to some tech and financial companies in order to help them secure their systems against its reported ability to exploit vulnerabilities.
\nBut that relies on those companies making sure their own access is tightly controlled.
\nThe person already had permission to view Anthropic’s AI models through work they had done for a third-party contractor, according to Bloomberg.
\nThe outlet also reported the group has been using the model since it gained access \u2013 although not for hacking, because they do not want to be detected.
\n“When powerful AI tools are accessed or used outside their intended controls, the risk is not just a security incident but the spread of capabilities that could be used for fraud, cyber abuse, or other malicious activity,” Saceanu said.<\/p>\n
In a speech to a large cyber-security conference on Wednesday, the head of the UK’s National Cyber Security Centre (NCSC) made a more positive case arguing AI tools can make things safer and more secure.
\nRichard Horne urged the CyberUK delegates not to fear new AI attacks, but to make sure they are doing the basics of cyber-security right.
\n“As we have seen in the media in recent days, frontier AI is rapidly enabling discovery and exploitation of existing vulnerabilities at scale<\/b>, illustrating how quickly it will expose where fundamentals of cyber-security are still to be addressed,” he said.
\nHorne’s warnings echo similar messages from recent years \u2013 for example the urgency for people to ensure they update the software on their systems and upgrade legacy IT.
\nAt the same event, Security Minister Dan Jarvis implored AI firms to work with the government on the “generational endeavour” to make sure AI is used to protect critical networks from attackers.
\nAll the most powerful and advanced AI models \u2013 known as frontier AI \u2013 are developed outside of the UK, with the top-tier companies based in the US or China.
\nThat means the UK relies on companies like Anthropic to give it access to Mythos and has no control over how it is built, trained or released.
\nOpenAI also has a cyber-security model it says is very capable called GPT 5.4 Cyber.
\nThe speeches at CyberUK also pressed home the ongoing threat of nation state and hacktivist attacks, particularly from Russia and China.
\nThe NCSC warns that cyber is now “the home front” of defence in the UK with recent events such as the Iran attacks showing that cyber plays an increasingly important role in all modern conflicts<\/b>.
\n#Anthropic #MythosAI #Cybersecurity #AIsecurity #UnauthorizedAccess #ClaudeMythos #FrontierAI #TechNews #DataProtection #CyberThreats<\/p>\n","protected":false},"excerpt":{"rendered":"
Anthropic is investigating a claim that a small group of people gained access to its Claude Mythos model \u2013 the cyber-security tool which the AI firm says is too powerful to release to the public. “We’re investigating a report claiming unauthorized access to Claude Mythos Preview through one of our third-party vendor environments,” the company […]<\/p>\n","protected":false},"author":1,"featured_media":4379,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[38],"tags":[],"class_list":["post-4378","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-world-news"],"_links":{"self":[{"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/posts\/4378","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/vanak.news\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4378"}],"version-history":[{"count":0,"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/posts\/4378\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/vanak.news\/index.php?rest_route=\/wp\/v2\/media\/4379"}],"wp:attachment":[{"href":"https:\/\/vanak.news\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4378"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/vanak.news\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4378"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/vanak.news\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4378"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}