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 Claude Mythos Preview 90%
- partners with AWS 80%
- used by Claude Code 70%
- used by OpenBSD 70%
- affiliated with Claude Mythos Preview 70%
- used by FreeBSD 70%
- founded Mythos 70%
- affiliated with Claude Opus 4.6 70%
- used by FFmpeg 70%
- partners with FFmpeg 70%
- affiliated with AI Security Institute 60%
2 day(s) with sentiment data
-
OpenAI launches Daybreak AI to automate cybersecurity threat detection
OpenAI has launched Daybreak, a new cybersecurity initiative designed to proactively identify and fix software vulnerabilities. This AI-driven program leverages specialized models like GPT-5.5-Cyber and the Codex Securi…
-
Google stops first AI-developed zero-day exploit
Google's Threat Intelligence Group has identified the first instance of cybercriminals using artificial intelligence to develop a zero-day exploit. This AI-generated tool was designed to bypass security measures in an o…
-
Project Glasswing revealed as deceptive AI 'kill switch'
Project Glasswing, purportedly a safety feature for AI models, is described as a 'kill switch' that is not what it seems. The name itself, referencing a butterfly with transparent wings, hints at a hidden or deceptive p…
-
Anthropic's withheld Mythos model signals urgent need for software quality
Anthropic has reportedly developed a powerful AI model named Claude Mythos, capable of discovering numerous zero-day vulnerabilities, including a long-standing bug in OpenBSD. The company decided against a public releas…
-
US government expands AI model pre-deployment review to Google, Microsoft, xAI
The U.S. government, through the Center for AI Standards and Innovation (CAISI), has expanded its pre-deployment evaluation of frontier AI models. New agreements with Google DeepMind, Microsoft, and xAI build upon exist…
-
US government secures pre-release AI model access for national security testing
The US Department of Commerce, through the Center for AI Standards and Innovation (CAISI), has expanded its pre-release access to advanced AI models from major tech companies for national security testing. Google DeepMi…
-
Project Glasswing launches $100M coalition for vulnerability intelligence, faces antitrust scrutiny
Project Glasswing has launched with a $100 million coalition and twelve initial partners, aiming to address vulnerability intelligence. The initiative introduces a tiered system for disclosing and patching security vuln…
-
EU lawmakers fail to agree on AI Act delays, Anthropic to testify on Mythos model
European Union legislators failed to reach an agreement on delaying key provisions of the AI Act, leaving the original August deadline for high-risk AI applications in place. Disagreements centered on whether products l…
-
Anthropic Launches New Security Tool for Enterprises
Anthropic has launched a public beta of its Claude Security tool for enterprise customers, designed to identify vulnerabilities within codebases and suggest patches. This tool operates using the Claude Opus 4.7 model an…
-
Anthropic's Claude Mythos Preview finds thousands of code vulnerabilities
Anthropic's Claude Mythos Preview model has identified thousands of critical code vulnerabilities across major operating systems and web browsers. To combat AI-assisted cyberattacks, Anthropic launched Project Glasswing…
-
Google Cloud Next 2026: AI Integration Dominates, Infrastructure and Security Take Center Stage
Google Cloud Next 2026 underscored the pervasive integration of AI across all technological domains. The event highlighted new AI-focused hardware, including specialized Tensor chips for training and inference, and a re…
-
Claude Mythos finds hundreds of Firefox bugs, sparking AI security debate
Anthropic's Claude Mythos model has demonstrated significant capabilities in identifying and helping to fix software vulnerabilities, including a 27-year-old bug in Firefox. While initial AI-generated bug reports were o…
-
Public AI models replicate Anthropic's vulnerability research findings
Vidoc Security has replicated findings from Anthropic's Mythos project using publicly available models like GPT-5.4 and Claude Opus 4.6. Their research indicates that advanced AI capabilities for identifying software vu…
-
Anthropic's Claude Opus 4.7 boosts coding and cybersecurity; OpenAI releases specialized models
Anthropic has released Claude Opus 4.7, which appears to be a significant advancement in coding capabilities and a smaller improvement in other areas. The company also introduced Claude Mythos, a model with advanced cyb…
-
Claude Mythos: China Reacts
Anthropic has announced Claude Mythos Preview, an AI model with advanced cybersecurity capabilities that emerged organically rather than through direct training. The model demonstrated its prowess by independently ident…
-
Anthropic's Mythos model finds decades-old code vulnerabilities, gains White House access
Anthropic's new model, Mythos, has demonstrated an unprecedented ability to discover long-standing vulnerabilities in open-source code that had evaded millions of automated tests and human experts for decades. This capa…
-
Anthropic's Claude Mythos targets critical software vulnerabilities via Project Glasswing
Anthropic is limiting the release of its powerful new model, Claude Mythos, due to its significant cybersecurity capabilities. The company is instead launching Project Glasswing, a program that will provide access to My…
-
Anthropic's Claude Mythos Preview highlights AI's systemic risk to markets
Anthropic's Claude Mythos Preview has significant implications for how systemic risk, particularly cyber risk, is priced. The ability of AI agents to autonomously construct exploits challenges the long-held assumption t…
-
Anthropic's Claude Mythos sparks debate over capabilities and cybersecurity risks
Anthropic has released details on its new Claude Mythos model, highlighting its advanced capabilities, particularly in cybersecurity, which has raised concerns about potential misuse. While the model demonstrates signif…
-
GPT-5.5 matches Anthropic's Mythos in cybersecurity tests
Anthropic's new Claude Mythos model, initially presented as a significant leap in cybersecurity capabilities, has been found to perform comparably to OpenAI's GPT-5.5 in recent tests. Researchers from the UK's AI Securi…