{"id":199782,"date":"2026-03-27T14:27:00","date_gmt":"2026-03-27T18:27:00","guid":{"rendered":"https:\/\/testing.news-you-need.com\/index.php\/2026\/03\/27\/anthropics-most-capable-ai-model-claude-mythos-leaks-deemed-major-cybersecurity-threat\/"},"modified":"2026-03-27T14:45:11","modified_gmt":"2026-03-27T18:45:11","slug":"anthropics-most-capable-ai-model-claude-mythos-leaks-deemed-major-cybersecurity-threat","status":"publish","type":"post","link":"https:\/\/testing.news-you-need.com\/index.php\/2026\/03\/27\/anthropics-most-capable-ai-model-claude-mythos-leaks-deemed-major-cybersecurity-threat\/","title":{"rendered":"Anthropic’s ‘Most Capable’ AI Model Claude Mythos Leaks, Deemed Major Cybersecurity Threat"},"content":{"rendered":"

Anthropic’s ‘Most Capable’ AI Model Claude Mythos Leaks, Deemed Major Cybersecurity Threat<\/a><\/p>\n

https:\/\/decrypt.co\/362606\/anthropic-most-capable-ai-model-claude-mythos-leaks<\/a><\/p>\n

Publish Date: 2026-03-27 14:27:00<\/a><\/p>\n

Source Domain: decrypt.co<\/a><\/p>\n

Author: <\/a><\/p>\n

Using an unordered list, summarize the following article with between 4 and 8 key points. In brief
\nA leaked draft post revealed Anthropic\u2019s most powerful AI model, Claude Mythos.
\nThe model also appears to introduce a new tier above Opus, internally referred to as \u201cCapybara.\u201d
\nCybersecurity stocks declined after reports suggested the system could accelerate AI-driven cyberattacks.
\nClaude creator Anthropic is developing a new AI model called Claude Mythos, described internally as the company\u2019s most capable model to date, with draft materials about the system being leaked online this week.The existence of the model was first reported by Fortune on Thursday after unpublished files tied to Anthropic\u2019s blog were discovered in a publicly accessible data cache. An Anthropic spokesperson confirmed the existence of the model to the publication.\u201cWe\u2019re developing a general purpose model with meaningful advances in reasoning, coding, and cybersecurity,\u201d an Anthropic spokesperson told Fortune. \u201cGiven the strength of its capabilities, we\u2019re being deliberate about how we release it. As is standard practice across the industry, we\u2019re working with a small group of early access customers to test the model. We consider this model a step change and the most capable we\u2019ve built to date.\u201dIn an archived development page reviewed by Decrypt, Anthropic called Mythos \u201cthe most powerful AI model we\u2019ve ever developed.\u201d\ufeff\u201cMythos is a new name for a new tier of model: larger and more intelligent than our Opus models\u2014which were, until now, our most powerful,\u201d Anthropic wrote. \u201cWe chose the name to evoke the deep connective tissues that link together knowledge and ideas.\u201dAccording to Anthropic, Mythos scored \u201cdramatically higher\u201d than Claude Opus 4.6 on tests of software coding, academic reasoning, and cybersecurity.The leak of Mythos appears to have originated from draft materials stored in an unsecured content management system. According to Fortune, Anthropic restricted public access to the data store after being notified that the files were searchable online. The company attributed the exposure to human error in the configuration of its CMS tools.However, Anthropic\u2019s documents labeled Mythos as version one of the new model, and described version two internally as \u201cCapybara,\u201d which the company also positioned above its current top-tier Opus models.The draft materials also highlighted concerns about the system\u2019s potential cybersecurity implications.\u201cAlthough Mythos is currently far ahead of any other AI model in cyber capabilities, it presages an upcoming wave of models that can exploit vulnerabilities in ways that far outpace the efforts of defenders,\u201d the company wrote.Because of those risks, the company said it plans to release the model cautiously, beginning with a limited early-access rollout aimed at organizations working on cybersecurity defense.Anthropic did not immediately respond to Decrypt\u2019s request for comment.While Anthropic took down the blog post, news of the leak quickly spilled into financial markets.Shares of several cybersecurity firms dropped after the reports surfaced, including Palo Alto Networks (PANW), which fell about 7%, and CrowdStrike (CRWD), which dropped roughly 6.4%. Meanwhile, Zscaler (ZS) declined around 5.8%, and Fortinet (FTNT)\u00a0slipped about 4% during Friday trading, according to Yahoo Finance.The selloff reaction echoes a similar market response to the reveal of a new Anthropic product. In February, Anthropic unveiled Claude Cowork, an AI system designed to automate complex workplace tasks\u2014including contract review and compliance\u2014which triggered a broad sell-off across software and professional-services companies.That sell-off erased roughly $285 billion in market value as investors reassessed the long-term impact of AI agents on enterprise software businesses.”The market’s response was a signal, not that AI agents will immediately replace these businesses, but that investors are finally pricing in the structural risk that foundation model providers can now compete directly with the software layer,” Nexatech Ventures founder Scott Dylan told Decrypt at the time. \u201cThat’s a polite way of saying if Anthropic can build a legal workflow tool in-house, what’s stopping them from doing the same for finance, procurement, or HR?\u201dDaily Debrief NewsletterStart every day with the top news stories right now, plus original features, a podcast, videos and more.
<\/p>\n","protected":false},"excerpt":{"rendered":"

Anthropic’s ‘Most Capable’ AI Model Claude Mythos Leaks, Deemed Major Cybersecurity Threat https:\/\/decrypt.co\/362606\/anthropic-most-capable-ai-model-claude-mythos-leaks Publish Date:…<\/p>\n","protected":false},"author":1,"featured_media":199783,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/cdn.decrypt.co\/resize\/1024\/height\/512\/wp-content\/uploads\/2025\/05\/ai-decrypt-style-05-gID_7.png","fifu_image_alt":"","footnotes":""},"categories":[15],"tags":[26,24,31],"class_list":["post-199782","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cybersecurity","tag-ai","tag-cybersecurity","tag-exploit"],"_links":{"self":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/199782"}],"collection":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/comments?post=199782"}],"version-history":[{"count":1,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/199782\/revisions"}],"predecessor-version":[{"id":199784,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/199782\/revisions\/199784"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media\/199783"}],"wp:attachment":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media?parent=199782"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/categories?post=199782"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/tags?post=199782"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}