Home / Technology / AI Uncovers 10,000+ Critical Software Flaws
AI Uncovers 10,000+ Critical Software Flaws
24 May
Summary
- AI initiative Project Glasswing has found over 10,000 vulnerabilities.
- These vulnerabilities are in essential software, identified by AI.
- Mythos Preview shows significant advancement in exploitation capabilities.
Launched last month, Anthropic's Project Glasswing has achieved a major milestone, identifying more than ten thousand high- or critical-severity vulnerabilities within essential software. This collaborative AI cybersecurity initiative involves Anthropic and its partners working together to enhance digital security.
Recent evaluations of Mythos Preview, an AI model developed by Anthropic, have shown remarkable progress. Notably, its exploitation capabilities have advanced significantly, marking a substantial leap compared to earlier AI models. These findings underscore the evolving power of AI in uncovering and potentially addressing complex security issues.
Disclaimer: This story has been auto-aggregated and auto-summarised by a computer program. This story has not been edited or created by the Feedzop team.
