Project Glasswing: Initial Update
Original title:Project Glasswing: An initial update
AI summary
Launched last month, Project Glasswing aims to use advanced AI models to secure critical software. Through about 50 partners using the Claude Mythos Preview model, it has discovered more than 10,000 high-severity or critical vulnerabilities in critical systems worldwide. Several partners reported more than a tenfold increase in vulnerability discovery efficiency. For example, Cloudflare found 2,000 vulnerabilities in critical-path systems; Mozilla discovered and fixed 271 vulnerabilities in Firefox 150, far more than previous-generation models. The project also scanned more than 1,000 open-source projects underpinning the internet, with independent verification accuracy reaching 90.6%. This marks a new stage in AI cybersecurity capabilities, with vulnerability discovery now far faster than manual efforts.
Key Takeaways
- Launched last month,Project Glasswing aims to use advanced AI models to secure critical software.
- Through about 50 partners using the Claude Mythos Preview model,it has discovered more than 10,000…
- Several partners reported more than a tenfold increase in vulnerability discovery efficiency.
- For example,Cloudflare found 2,000 vulnerabilities in critical-path systems
- Mozilla discovered and fixed 271 vulnerabilities in Firefox 150,far more than previous-generation…
Quick Facts
- Published
- Updated
- Source
- Anthropic:Newsroom(网页)
- Category
- Industry
- AI score
- 90
Mythos Preview用一个月找到的漏洞比过去一年还多,网络安全从「发现难」变成「修不过来」,这个转折点所有安全从业者都得读。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Project Glasswing: Initial Update. May 22, 2026. https://new.nat6.net/en/news/cmphcciiv0n76sljw2c5qbnwe/