Anthropic Expands Project Glasswing to 150 Orgs After Partners Found 10,000+ Flaws

ConradBastable · x · 2026-09-04

• Anthropic is expanding Project Glasswing, granting roughly 150 new organizations across 15+ countries access to Claude Mythos Preview, adding power, water, healthcare, communications, and hardware sectors plus widely relied-upon code maintainers.
• The initial 50 partners have already used the model to scan codebases, uncovering more than 10,000 high- or critical-severity vulnerabilities; Anthropic estimates a major attack on most partners could affect over 100 million people.
• The move follows collaboration with security firms, open-source maintainers, and the US government, aiming to make all software more secure as AI reshapes cybersecurity assumptions. A tweet accompanying the post claims Anthropic gave preselected big-tech friends early access to Mythos since April.

Original post →

More from Models

• OpenAI's Astra can now layout and route PCBs, sparking hardware engineering debate — MikePFrank · 2026-09-04
• Ethan Mollick: Astra just takes action, spinning up agents on vague requests — emollick · 2026-09-04
• 'AGI is 74% deepswe': GPT-6-Astra benchmark results become an AI-circle meme — amaarora · 2026-09-04
• Grok 4.7 reportedly days away, trained on SpaceX engineering data; Grok 4.6 already ties GPT-6 Astra at 61 — XFreeze · 2026-09-04
• Ex-OpenAI safety lead Miles Brundage: if your primary emotion on AI isn't concern, you're misreading it — Miles_Brundage · 2026-09-04
• Gary Marcus on GPT-6 Astra: symbolic world models are vindication, but no proof of AGI — GaryMarcus · 2026-09-04

Models channel →