Anthropic has launched Project Glasswing, a global cybersecurity effort. The initiative brings together major technology and infrastructure players to tackle the growing threat of AI-driven cyberattacks. The partners in this program are Amazon Web Services, Apple, Google, Microsoft, NVIDIA, and the Linux Foundation.
The goal is to deploy advanced AI systems that can detect and fix software vulnerabilities before attackers exploit them.
At the center of the initiative sits an unreleased AI system, Claude Mythos Preview. Anthropic claims that the model can detect and exploit software vulnerabilities at a level that surpasses all but the most skilled human experts.
From this, it was found that there were more than 2000 significant vulnerabilities in standard operating systems and internet browsers. This shows the rapid evolution of artificial intelligence technology, as well as the threats that have arisen alongside this rapid growth.
Anthropic has shown concern regarding how soon such technologies might be put to ill use. According to the scientists behind the study, the same technologies that boost cybersecurity can be used to carry out high-tech attacks. The idea is to remain ahead of the game by deploying such AI technologies as a defense mechanism.
Also Read: OpenAI, Google, Anthropic Unite to Counter AI Model Replication & Reshape Global Tech Race
More than 40 organizations involved in building and maintaining critical software infrastructure have received access to Mythos Preview. Partner companies will use the system to scan their platforms, identify weaknesses, and share insights across the ecosystem.
Moreover, Anthropic is ready to contribute $100 million in usage credits, in addition to funding initiatives to secure open-source code.
According to Anthropic, cybersecurity problems require joint action. Anthropic is collaborating with US government representatives to highlight the essential role of cyber-infrastructure safety in national security in the age of artificial intelligence.