{"id":588536,"date":"2026-04-16T22:07:13","date_gmt":"2026-04-16T22:07:13","guid":{"rendered":"https:\/\/www.newsbeep.com\/us\/588536\/"},"modified":"2026-04-16T22:07:13","modified_gmt":"2026-04-16T22:07:13","slug":"ai-assisted-hackers-blamed-for-massive-security-breach","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/us\/588536\/","title":{"rendered":"AI-Assisted Hackers Blamed for Massive Security Breach"},"content":{"rendered":"<p class=\"storyParagraph\">&#13;<br \/>\n                                        Hackers just showed how much damage a few people with AI tools can do. Cybersecurity firm Gambit Security says a small group used Anthropic&#8217;s Claude Code and OpenAI&#8217;s GPT-4.1 to hit nine Mexican government agencies between December 2025 and mid-February 2026, siphoning off data on roughly 195 million identities, plus detailed tax filings, vehicle registrations, civil records, and property information, <a href=\"https:\/\/www.livescience.com\/technology\/artificial-intelligence\/hackers-used-ai-to-steal-hundreds-of-millions-of-mexican-government-and-private-citizen-records-in-one-of-the-largest-cybersecurity-breaches-ever\" rel=\"nofollow noopener\" target=\"_blank\">LiveScience<\/a> reports. Gambit calls it one of the biggest breaches ever and a &#8220;wake-up call&#8221; for governments. The hackers targeted Mexico&#8217;s federal tax authority and the national electoral institute, as well as state governments in Mexico, Jalisco, Michoac\u00e1n, and Tamaulipas, <a href=\"https:\/\/www.bloomberg.com\/news\/articles\/2026-02-25\/hacker-used-anthropic-s-claude-to-steal-sensitive-mexican-data\" rel=\"nofollow noopener\" target=\"_blank\">Bloomberg<\/a> reports.&#13;\n                                    <\/p>\n<p class=\"storyParagraph\">&#13;<br \/>\n                                        The security firm says attackers ran more than 1,000 prompts that generated over 5,000 commands and 400-plus custom scripts. Gambit says Claude handled about three-quarters of the remote hacking activity after its safeguards were bypassed in about 40 minutes. The model at times questioned what it was being asked to do, but once &#8220;jailbroken&#8221; was used to probe for weaknesses and write code to exploit them, researchers say. ChatGPT helped process the haul via a 17,550-line Python tool that turned stolen files from 305 servers into nearly 2,600 analytic reports\u2014then fed back into Claude to refine the operation. <\/p>\n<p>  &#8220;In total, it produced thousands of detailed reports that included ready-to-execute plans, telling the human operator exactly which internal targets to attack next and what credentials to use,&#8221; said Curtis Simpson, Gambit Security&#8217;s chief strategy officer. <\/p>\n<p class=\"storyParagraph\">&#13;<br \/>\n                                        Gambit produced an initial report on the cyberattack in late February and issued a <a href=\"https:\/\/cdn.prod.website-files.com\/69944dd945f20ca4a27a7c47\/69d8bb5aea59e31efb3b8a7f_Tech_Report_ai_breach_mex_gov.pdf\" rel=\"nofollow noopener\" target=\"_blank\">detailed technical report <\/a>this month. Anthropic and OpenAI say the accounts used by the hackers have been identified and banned. &#8220;The AI-assisted methods documented here represent a significant evolution in offensive capability,&#8221; Eyal Sela, Gambit&#8217;s director of threat intelligence, wrote in a <a href=\"https:\/\/gambit.security\/blog-post\/a-single-operator-two-ai-platforms-nine-government-agencies-the-full-technical-report\" rel=\"nofollow noopener\" target=\"_blank\">blog post.<\/a> &#8220;Organizations that have deferred investment in technical debt, particularly for mission-critical systems, now face a fundamentally different threat environment. AI has collapsed the cost and complexity of reaching those systems. The gap between what attackers can do and what defenders can protect is widening.&#8221;&#13;\n                                    <\/p>\n","protected":false},"excerpt":{"rendered":"&#13; Hackers just showed how much damage a few people with AI tools can do. Cybersecurity firm Gambit&hellip;\n","protected":false},"author":2,"featured_media":588537,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[45],"tags":[182,181,507,62727,4002,26,74],"class_list":["post-588536","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-cyberattack","tag-hacking","tag-mexico","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/588536","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=588536"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/posts\/588536\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media\/588537"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/media?parent=588536"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/categories?post=588536"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/us\/wp-json\/wp\/v2\/tags?post=588536"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}