Anthropic launches Project Glasswing, an effort to prevent AI cyberattacks with AI
We see a lot of pessimism about the potential negative impacts of artificial intelligence, particularly regarding how it could create new cybersecurity challenges. Anthropic announced a new initiative called Project Glasswing to help address these concerns by working “to secure the world’s most critical software” against AI-based attacks. This project includes Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks as partners.
Participants will use Claude Mythos Preview, a first-of-its-kind multi-purpose template from Anthropic, to enhance their own security projects. Anthropic claims that this model discovered thousands of exploitable vulnerabilities, “including some across all major operating systems and web browsers.” The company said it wants to start using its tools defensively to prevent malicious use of AI that could have serious economic and security consequences.
Anthropic has become one of the most notable AI companies raising concerns about ethics in this area. Earlier this year, the company refused to remove guardrails on its services intended for use by the Pentagon, prompting the Department of Defense to sanction Anthropic with a “supply chain risk” designation in retaliation. The launch of Project Glasswing could be a useful start toward improving cybersecurity in the AI age, but some damage has already been done. His own Claude was allegedly used by a hacker against several government agencies in Mexico in February.
