An initiative to secure the world's software | Project Glasswing
- Source
- youtube.com
- Author
- Anthropic
- Date
Context
Project Glasswing is Anthropic's effort to put a new AI model's vulnerability-finding ability into the hands of software defenders. Anthropic says its unreleased Claude Mythos Preview can surpass all but the most skilled humans at finding and exploiting software vulnerabilities, a company claim about its own model. Launch partners, including Amazon Web Services, Apple, Cisco, Google, Microsoft, and the Linux Foundation, will use it in defensive security work, and Anthropic says over 40 additional organizations received access to scan first-party and open-source systems. Program announcement
The reasoning is that the same ability could serve attackers or defenders. Anthropic warns that such capabilities will proliferate before long, calls the project "an urgent attempt to put these capabilities to work for defensive purposes," and says it will share what it learns with the industry. Whether defenders actually stay ahead is the hope behind the program, not something the announcement demonstrates. Program rationale
Checking sign-in…
Loading comments…






