Anthropic has unveiled a preview of a new artificial intelligence (AI) model called Mythos that it says is the most powerful model ever developed. Mythos is public to the Claude ecosystem, with strong coding and reasoning skills.
But it also has the ability to scan software for vulnerabilities and code pad vulnerabilities. Athropic says that in internal testing, Mythos has found thousands of zero-day vulnerabilities, including those 1–2 decades old. Because of this capability and its impact on security, Mythos is not open to the public.
Instead, the preview for Mythos is only available through Project Glasswing, a cybersecurity initiative involving 11 partner organizations Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. In addition to these 11 partners, 40 additional organizations will also be given access to Project Glasswing.
The results of using the model will then be shared for the benefit of the broader cybersecurity industry.
