AI systems becoming attack surfaces and attack tools simultaneously

— not a single breach but a phase transition. The same week brings: AI assistants weaponised as C2 relays, AI-assisted mass compromises of network infrastructure, supply chain attacks on AI coding tools, and AI models deployed to design drugs entering advanced human trials. Google announces AI that transforms Earth surface analysis for environmental monitoring; Anthropic rolls out Claude Code Security to patch vulnerabilities; check-marked AI systems scan blood cells better than human experts. ...