Guardrails Breached: The New Reality of GenAI-Driven Attacks

The article aggregates recent media and vendor reports documenting an accelerating trend: threat actors are leveraging generative AI for more persuasive social engineering ("vibe hacking"), automating malware development (including AI-generated RATs and polymorphic code), weaponizing fake AI websites to distribute infostealers, and exploiting high-severity vulnerabilities such as EchoLeak (CVE-2025-32711). It warns that APTs and cybercriminals are experimenting with and adopting GenAI tools faster than defenders, creating new attack surface and operational scale.