# Project Glasswing Type: initiative Source: GetYourBrief — https://getyourbrief.com/entity/project-glasswing Canonical HTML page: https://getyourbrief.com/entity/project-glasswing ## Timeline - **2026-06-24**: Anonymous Official Confirms Details — A U.S. official confirms to the Associated Press that Mythos identified vulnerabilities within hours but did not necessarily exploit them. - **2026-06-24**: Zhou Hongyi warns China needs its own Mythos — At a cybersecurity conference in Beijing, 360 founder Zhou Hongyi calls Mythos a 'cyber nuclear weapon' and demands China develop a domestic equivalent to achieve reciprocal strategic deterrence. - **2026-06-13**: Anthropic pledges further details — Anthropic states it will share additional information about the jailbreak and the directive within the next 24 hours. - **2026-06-12T17:21:00-04:00**: Export control directive received — At 5:21 p.m. ET, the U.S. government issues an export control directive ordering Anthropic to suspend all access to Fable 5 and Mythos 5 for any foreign national, including foreign national employees. - **2026-06-12T17:21:00-04:00**: Government Directive Received — Anthropic receives an export control directive from the US government at 5:21 PM ET ordering suspension of all access to Fable 5 and Mythos 5 by any foreign national, including foreign national Anthropic employees. - **2026-06-12**: Models disabled — Anthropic abruptly disables Fable 5 and Mythos 5 for all customers to ensure compliance with the directive. - **2026-06-12**: Access Disabled for All Customers — Anthropic abruptly disables Fable 5 and Mythos 5 for all customers worldwide to ensure compliance with the directive, while confirming all other Anthropic models remain unaffected. - **2026-06-11**: Senate Hearing Disclosure — Senator Mark Warner reveals the Mythos testing results during a Senate Banking Committee hearing, stating the tool broke into almost all classified systems in hours. - **2026-06-09**: Fable 5 and Mythos 5 Announced — Anthropic announces Fable 5 and Mythos 5, with Fable 5 marking the company's first release of a highly advanced model to the public, featuring new safeguards that block responses in specific high-risk areas. - **2026-06-08**: Announcement of Fable 5 and Mythos 5 — Anthropic publicly unveils Fable 5 and Mythos 5, with Fable 5 marking the first such advanced model made available to the public using new safeguards. - **2026-06**: Administration Directive Issued — The Trump administration restricts the use of some Anthropic models, escalating tensions over AI military deployment. - **2026-04**: Limited release of Claude Mythos Preview — Anthropic releases Claude Mythos Preview with advanced cybersecurity capabilities under a restricted initiative called Project Glasswing, catching the attention of Wall Street and government officials. - **2026-04**: Claude Mythos Preview Released — Anthropic releases Claude Mythos Preview with advanced cybersecurity capabilities, stunning Wall Street and government officials. Access limited to select companies under Project Glasswing. - **2026-04**: Anthropic releases Mythos and launches Project Glasswing — Mythos is released with autonomous vulnerability discovery capabilities, and Project Glasswing grants more than 40 US-allied organizations access to Mythos Preview. China is excluded from the alliance. - **2026**: Project Glasswing Testing Initiated — Anthropic and U.S. intelligence agencies conduct a testing exercise where Mythos identifies vulnerabilities in classified systems within hours. ## Recent coverage (10 stories, network-wide) ### Mythos AI turned hours into zero-day discovery for US classified systems 2026-07-12 11:12:47 · Sector: cyber · Sentiment: Negative · Impact: 8/10 · Sources: 2 An AI red-teaming exercise using Anthropic’s Mythos model identified vulnerabilities across almost all classified U.S. government networks in hours, compressing the traditional weeks-long security audit timetable. The finding points to a future where autonomous vulnerability scanners could dominate cyber defense and offense. Full story: https://getcyberbrief.com/story/mythos-ai-cyber-vulnerability-discovery-speed ### 100x faster vulnerability finding: China’s defenders face ‘cyber nuclear gap’ 2026-06-26 00:53:03 · Sector: cyber · Sentiment: Negative · Impact: 8/10 · Sources: 4 Anthropic’s Mythos AI makes vulnerability discovery 100x faster and cheaper, prompting 360 founder Zhou Hongyi to warn that China’s exclusion from the Project Glasswing alliance leaves its digital infrastructure dangerously exposed. He calls for a homegrown equivalent to restore strategic balance. Full story: https://getcyberbrief.com/story/china-mythos-cyber-nuclear-gap ### AI that finds bugs 100x faster: 360 founder demands Chinese Mythos 2026-06-26 00:52:53 · Sector: ai · Sentiment: Negative · Impact: 8/10 · Sources: 4 Anthropic’s release of Mythos, an AI model that autonomously uncovers software vulnerabilities at 100 times the speed of human researchers, has triggered a call from 360’s Zhou Hongyi for China to build its own frontier cybersecurity AI. The gap risks a new AI arms race with direct national security implications. Full story: https://getaibrief.com/story/china-ai-race-mythos-competitor ### Anthropic's Mythos cracked classified systems in hours, risking IPO and export ban 2026-06-24 06:23:19 · Sector: startups · Sentiment: Negative · Impact: 9/10 · Sources: 5 Anthropic's Mythos AI found vulnerabilities in classified U.S. systems within hours, but the breakthrough comes as the startup faces an export ban and blacklisting by the Trump administration, threatening its IPO and growth prospects. Full story: https://getstartupbrief.com/story/anthropic-mythos-classified-vulnerabilities-startup-ipo-risk ### Mythos AI cracks US classified systems in hours, raising AI alignment crisis 2026-06-24 06:23:08 · Sector: ai · Sentiment: Negative · Impact: 9/10 · Sources: 5 Anthropic's Mythos model demonstrated the ability to find vulnerabilities in almost all classified U.S. government networks in hours, intensifying the debate over AI safety, dual-use risks, and the future of automated cyber offense. Full story: https://getaibrief.com/story/mythos-ai-classified-vulnerabilities-alignment-safety ### Mythos AI Breached Classified Systems in Hours, Raising Space Defense Alarms 2026-06-24 03:35:10 · Sector: space · Sentiment: Negative · Impact: 9/10 · Sources: 2 Anthropic’s Mythos AI model found vulnerabilities in classified U.S. systems within hours during a Project Glasswing test, according to a U.S. official and Senator Mark Warner. The revelation poses critical risks for space-based assets, missile warning systems, and nuclear command and control networks that rely on these classified infrastructures. The incident intensifies the race to harden defense networks with AI-driven countermeasures. Full story: https://getspacebrief.com/story/mythos-ai-classified-breach-space-defense ### Mythos AI Uncovers Classified System Flaws in Hours, Sparking Cyber Defense Race 2026-06-24 03:34:59 · Sector: cyber · Sentiment: Negative · Impact: 9/10 · Sources: 2 A testing exercise revealed Anthropic’s Mythos model can identify vulnerabilities inside classified U.S. systems in hours, a capability that reshapes the cybersecurity landscape. While the model reportedly did not exploit the flaws, the speed of discovery accelerates the imperative for AI-driven patch management and zero-trust architectures. The incident may also drive new regulatory mandates for AI red-teaming in federal systems. Full story: https://getcyberbrief.com/story/mythos-ai-classified-vulnerability-discovery ### Export Control Directive Kills Access to Fable 5 and Mythos 5 Over a Single Jailbreak Claim 2026-06-13 04:39:43 · Sector: ai · Sentiment: Negative · Impact: 8/10 · Sources: 4 The US government's export control directive against Anthropic's Fable 5 and Mythos 5 hinges on a single, narrow jailbreak technique that the company says reveals only minor, previously known vulnerabilities — a capability it claims is already present in OpenAI's GPT-5.5 and other public models. The decision to classify AI model access as an export-controlled item signals a paradigm shift in AI governance that could reshape the entire frontier model landscape. Full story: https://getaibrief.com/story/anthropic-fable-5-mythos-5-export-control-jailbreak ### US Government Orders Sudden Shutdown of 2 AI Models Serving Hundreds of Millions 2026-06-13 04:39:05 · Sector: saas · Sentiment: Negative · Impact: 8/10 · Sources: 4 The US government's abrupt directive forcing Anthropic to disable Fable 5 and Mythos 5 exposes a critical vulnerability in the SaaS ecosystem: cloud-hosted AI models can be terminated by regulatory fiat with zero notice, leaving enterprise customers stranded. The export control order specifically targets foreign national access, raising urgent compliance questions for any SaaS provider serving global customers with AI-powered features. Full story: https://getsaasbrief.com/story/anthropic-fable-mythos-export-control-saas-risk ### US directive forces Anthropic to disable 2 AI models, raising due process alarms 2026-06-13 01:36:31 · Sector: legal · Sentiment: Negative · Impact: 8/10 · Sources: 3 Anthropic’s abrupt disabling of Fable 5 and Mythos 5 under a US export control order lacking detailed justification highlights executive overreach and transparency gaps in AI regulation. Legal experts warn of insufficient statutory process and opaque national security claims. Full story: https://getlegalbrief.com/story/export-control-anthropic-fable-mythos-due-process --- This page is a machine-readable summary. Sentiment measures the directional read of each development for this entity, not the tone of the reporting; impact weights consequence, not syndication reach. See https://getyourbrief.com/guides/methodology for the full editorial methodology.