This Wednesday morning, Anthropic's AI, Mythos, is uncovering significant security flaws in Apple's macOS, an exploit that could grant an unprivileged local user complete device access. Both Mashable and ETF Database confirm this AI is so effective it's not publicly available, with Anthropic committing $100 million in usage credits to its Project Glasswing initiative.

This rapid discovery highlights a critical shift: the shrinking timeline between vulnerability discovery and exploitation. The I