Lyptus Research finds that scaling laws increase cyberattack potency by 10×. The study shows larger models can generate more sophisticated phishing templates and exploit code. Import AI highlights the trend, warning that defenders must adopt AI‑driven detection. Practitioners should prioritize model‑aware threat intelligence and rapid patching, and continuous monitoring to stay ahead.