Anthropic Launches Glasswing, Uses Claude Mythos AI
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 employ its Claude Mythos Preview model to locate and remediate software vulnerabilities.
- Claude Mythos Preview is described by multiple outlets as a highly powerful AI model capable of uncovering thousands of previously unknown zero‑day bugs across major operating systems and browsers.
- Project Glasswing brings together partners Amazon Web Services, Apple, Broadcom, Nvidia, and Microsoft to jointly apply the model for software security.
- The Register ran the headline 'All your zero‑days are belong to Mythos', underscoring the model’s extensive vulnerability detection.
- Wired highlighted that the model’s ability to find thousands of zero‑day vulnerabilities is being harnessed in Project Glasswing, emphasizing both its power and the need for restricted access.
Why it matters: The partnership of Amazon Web Services, Apple, Broadcom, Nvidia and Microsoft means the defensive capability can be deployed across a broad ecosystem, potentially reducing the window of exposure for critical software, whereas the model’s power and restricted access highlight the tension between security gains and the risk of weaponization.
Ask SkimNews
