Anthropic, the company behind the Claude family of models, is stepping into a more critical role in the artificial intelligence (AI) race – one that goes beyond building intelligence to securing it. Its latest initiative, Project Glasswing, is a cybersecurity-focused effort powered by its most advanced model, Claude Mythos Preview, built to detect and even exploit software vulnerabilities at a level that rivals top human experts. The goal is to use AI offensively to strengthen defenses before bad actors can catch up.
Glasswing is not a solo play. It brings together a deep bench of partners, including Microsoft Corporation (MSFT), Broadcom (AVGO), and NVIDIA Corporation (NVDA), alongside other cybersecurity leaders and tech giants. The model has already uncovered thousands of high-severity flaws across major systems, highlighting both the risks and urgency behind this push. Anthropic is committing up to $100 million in credits, with access limited to select participants through platforms like cloud and API integrations.
