🧰 Project Glasswing

The items that name it 19

• Was I wrong about open models & cyber risks?
  articleas tools2026-10-30

  Anthropic announcement (April 7, 2026) that warned such capabilities could proliferate with severe fallout for economies, public safety and national security

• LLMs are making formal software verification feasible at a much larger scale
  articleas classification.key entities2026-10-22
• AI Is Calling In Every Risk You've Accepted: Seven Insights From The AI Security Panel: Unfiltered
  articleas classification.key entities2026-10-21
• Anthropic Gives Open Source Maintainers a Fast Track to Mythos Vulnerability Findings
  articleas classification.key entities2026-10-15
• Anthropic Fast-Tracks AI Bug Reports to OSS Maintainers, Taps 11 Firms for OT Security
  articleas tools2026-10-09

  Anthropic program whose partners uncovered vulnerabilities; lessons from it informed the two new initiatives

• Anthropic Introduces 3-Tier Cyber Verification Program for AI Access
  articleas tools2026-10-07

  Anthropic program that gave organizations securing critical software access to Claude Mythos. It is now merged into the Cyber Verification Program, and existing members are moving to the Specialized Access tier. Partners found at least 129,000 verified vulnerabilities between April and July.

• Expanding the Cyber Verification Program
  articleas tools2026-10-06

  Earlier program that gave organizations securing the most critical software access to Claude Mythos. Partners found at least 129,000 verified software vulnerabilities between April and July 2026.

• CVE-2026-61500: Rejetto HFS Session Forgery Vulnerability (found via Anthropic Mythos)
  writeupas classification.key entities2026-09-30
• GLM-5.3 and the spread of advanced cyber capabilities
  articleas tools2026-09-29

  Program enabling trusted cyber defenders to use Claude's advanced cyber capabilities to find vulnerabilities (over 10,000 found) ahead of malicious actors

• 2026 in LLMs (so far)
  talkas classification.key entities2026-09-27
• Rumour Is the Exploit — AI agents collapse time-to-exploit to near-zero
  articleas tools2026-08-22It is written here as Project Glasswing (frontier model access program, expanded to 150 organisations across 15 countries).

  frontier model access program, expanded to 150 organisations across 15 countries

• Visa Vulnerability Agentic Harness (VVAH)
  repoas tools2026-07-29

  Anthropic's initiative for AI-assisted vulnerability research; VVAH is built on learnings from it (white paper cited as technical background)

• The OpenSourceMalware Show #12
  talkas classification.key entities2026-07-09
• Full Fathom Five: The context of Anthropic’s Mythos-class public release
  articleas classification.key entities2026-06-13
• Move over, Mythos. Here comes... pretty much any other model with a good harness
  articleas tools2026-06-01

  Referenced alongside Mythos as an alternative to relying on a single frontier model for vulnerability discovery.

• Project Glasswing: Initial Update (Anthropic)
  articleas classification.key entities2026-05-27
• AI-Assisted Vulnerability Discovery
  articleas tools2026-05-15

  Anthropic coalition program funneling access to AI vulnerability discovery models to partners including AWS, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorgan Chase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks

• Mythos AI Security Testing
  articleas classification.key entities2026-05-12
• AWS, Google, and Vercel Agent Flaws Let Attackers Trigger Tools Without Running the Model
  articleas classification.key entities