Anthropic Launches Glasswing AI to Spot Zero‑Day Flaws
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic announced Project Glasswing, a cybersecurity initiative that will use its Claude Mythos Preview model to locate and fix software vulnerabilities.
- Project Glasswing unites Amazon Web Services, Apple, and Broadcom as collaborative partners in the effort.
- Claude Mythos Preview is the AI model at the core of the initiative, touted for its ability to uncover zero‑day flaws across major operating systems and browsers.
- The Register reported that Anthropic’s model identifies thousands of zero‑day vulnerabilities, highlighting its unprecedented detection capability.
- Shelly Palmer reported that Anthropic deemed the Claude Mythos Preview model too dangerous for public release, underscoring concerns over its misuse.
Why it matters: Anthropic’s partnership with AWS, Apple, and Broadcom equips software vendors and security teams with a powerful AI tool to discover and patch zero‑day flaws, potentially reducing the $21 billion cybercrime losses reported by the FBI. However, the decision to keep the Claude Mythos Preview model private limits broader industry access and may concentrate defensive capabilities within the participating firms.
Ask SkimNews
