
Attackers are already using AI to research targets, improve language and adapt social engineering at speed. The 2026 SANS Security Awareness & Culture Report shows that 58% of awareness teams have not moved beyond first experiments, and most of the rest use AI to produce content rather than to change behavior.










