Project Glasswing
PulseAugur coverage of Project Glasswing — every cluster mentioning Project Glasswing across labs, papers, and developer communities, ranked by signal.
- developed by Anthropic 100%
- developed by Mythos 95%
- partners with Claude Mythos Preview 90%
- developed by Claude Mythos Preview 90%
- uses Claude Mythos Preview 90%
- used by Mythos 90%
- used by Claude Opus 4.8 90%
- used by OpenBSD 90%
- instance of Claude Opus 4.8 90%
- instance of Claude Mythos 5 90%
- instance of Claude Opus-4.6 90%
- used by Mythos Preview 90%
- 2026-06-05 research_milestone Anthropic's Project Glasswing AI autonomously discovered thousands of high-severity vulnerabilities. source
- 2026-06-05 product_launch Anthropic has launched Project Glasswing, a new initiative focused on enhancing cybersecurity through AI. source
- 2026-06-03 product_launch Anthropic expanded its Project Glasswing initiative, which uses the Claude Mythos AI model to identify and fix software vulnerabilities, to include 150 organizations across 15 countries. source
- 2026-06-02 product_launch Anthropic is scaling up Project Glasswing with 150 partners across more than 15 countries to scan critical infrastructure for security flaws. source
- 2026-06-02 product_launch Anthropic is expanding Project Glasswing to include approximately 200 organizations. source
- 2026-06-02 product_launch Anthropic is expanding Project Glasswing to approximately 200 vetted organizations across over 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding its Project Glasswing initiative to 150 organizations across 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding Project Glasswing to include approximately 200 organizations across over 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding its AI-powered Project Glasswing initiative to include critical infrastructure organizations across 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding its AI-powered vulnerability scanning initiative, Project Glasswing, to include critical infrastructure organizations in over 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding its Project Glasswing preview program to over 150 additional companies in more than 15 countries. source
- 2026-06-02 product_launch Anthropic is expanding its Project Glasswing initiative to include approximately 150 new organizations. source
- 2026-05-23 product_launch Anthropic released initial findings from its Project Glasswing AI initiative.
- 2026-05-22 research_milestone Anthropic's Project Glasswing initiative has identified over ten thousand high- or critical-severity vulnerabilities in essential software.
- 2026-05-22 product_launch Anthropic launched Project Glasswing, a cybersecurity initiative using AI to find software vulnerabilities.
20 day(s) with sentiment data
-
Anthropic report details AI misuse shift to agentic attacks
Anthropic has released a report detailing how malicious actors misuse AI models, particularly focusing on the shift from simple malware writing to more sophisticated agentic actions like lateral movement within networks…
-
China needs its own 'cyber nuclear weapon' AI model, says 360 founder
Zhou Hongyi, founder of 360 Security Technology, has urged China to develop its own AI model equivalent to Anthropic's Mythos. He described Mythos as a
-
Anthropic's restricted AI model Mythos still accessible to some early testers
Despite a U.S. government order to halt distribution of its powerful AI model, Claude Mythos, Anthropic appears to be allowing some early testers within its "Project Glasswing" initiative to retain access. The U.S. gove…
-
Anthropic's Mythos Preview remains accessible to ~200 partners post-US order
Approximately 200 organizations participating in Anthropic's Project Glasswing continue to have access to the Mythos Preview model. This access persists despite a recent US government order that restricted broader avail…
-
AI vulnerability discovery to favor defense after bumpy 2026-2027 transition
AI models are poised to significantly enhance vulnerability discovery, shifting the landscape from sparse sampling to dense sampling. This transition, while ultimately favoring defense, is expected to cause a bumpy peri…
-
Anthropic disables AI models over geopolitical access concerns · 1 source tracked
Anthropic has disabled its Claude "Mythos" and Fable 5 AI models due to geopolitical concerns, specifically regarding SK Telecom's access and alleged links to China. The White House reportedly ordered Anthropic to revok…
-
US restricts Anthropic AI models over SK Telecom-China concerns · 6 sources tracked
The Trump administration has imposed export controls on Anthropic's advanced AI models, including Claude Mythos and Fable 5, due to concerns over SK Telecom's access. US officials suspected SK Telecom, a South Korean te…
-
US restricts Anthropic AI models over security fears; experts warn of wider trend
Anthropic has temporarily taken its Claude Fable 5 and Mythos 5 AI models offline following a directive from the U.S. government. The directive restricts foreign nationals from using the services, citing national securi…
-
Anthropic's 'Too Capable' Claude Mythos AI Faces Release Dilemma
Anthropic has developed an advanced AI model named Claude Mythos, also referred to as Project Glasswing, which is reportedly too capable for public release. The company is grappling with the ethical implications and pot…
-
Anthropic's Claude Fable 5 launches with safety guardrails for public access
Anthropic has launched Claude Fable 5, a new model positioned as safe for broad public access, with safeguards designed to route sensitive queries to a more restricted model, Claude Opus 4.8. The company claims these sa…
-
Anthropic suspends Fable 5 and Mythos 5 access due to US government order
Anthropic has temporarily suspended customer access to its Fable 5 and Mythos 5 AI models following a directive from the US government. The government cited national security concerns, potentially related to a discovere…
-
Anthropic's Claude Fable 5 routes queries to Opus 4.8 via classifiers
Anthropic has launched Claude Fable 5, which utilizes a novel architecture where specific queries are routed to Claude Opus 4.8. This routing mechanism, triggered by classifiers detecting sensitive topics like cybersecu…
-
Cohere releases North Mini Code for developer AI control
Cohere has released North Mini Code, an open-source, mixture-of-experts model designed for developers seeking greater control and transparency over their AI technology. This 30-billion parameter model is available under…
-
Anthropic's Claude Fable 5 to Revolutionize Cybersecurity
Anthropic has released Claude Fable 5, a new model that is expected to significantly impact the cybersecurity landscape. This model, previously known as Project Glasswing, was developed with a focus on security applicat…
-
LessWrong questions Anthropic's AI pause motivations
A LessWrong post questions the motivations behind Anthropic's recent article on recursive self-improvement (RSI) and their proposal for a pause in AI development. The author argues that while Anthropic may benefit finan…
-
Anthropic ships dual-model Claude Fable 5 with advanced coding and safety features
Anthropic has released Claude Fable 5, a model that the company deems too dangerous for unrestricted release. The model is a dual system: a public-facing version, Fable 5, uses a classifier to route potentially risky qu…
-
Anthropic prices Claude Fable 5 at premium, limits sensitive queries
Anthropic has released its most advanced model, Claude Fable 5, with a premium pricing structure and limited access. The model is priced at double the rate of its predecessor and free access will be removed from paid su…
-
Anthropic releases Claude Fable 5, its most powerful public AI model
Anthropic has released its latest AI models, Claude Fable 5 and Claude Mythos 5. Fable 5 is positioned as their most capable publicly available model for general users, while Mythos 5 is being offered to select cybersec…
-
Anthropic launches Fable 5 and Mythos 5 models with advanced reasoning
Anthropic has released two new models, Fable 5 and Mythos 5, with a strong emphasis on reasoning capabilities, particularly for cybersecurity applications. Fable 5, a general-use model, demonstrates advanced performance…
-
Anthropic releases dual AI models: Mythos 5 for partners, Fable 5 with safety guardrails
Anthropic has released two new AI models, Claude Fable 5 and Claude Mythos 5. The Mythos 5 model, which possesses advanced capabilities including potential exploitation for cybersecurity threats, is being offered only t…