What happened
RAND published a 23-page Expert Insights paper arguing that agentic AI — systems capable of pursuing goals with a degree of autonomy — will 'alleviate the constraint on expert-level human capacity' that has historically confined cyber operations to tactical rather than strategic effects. Author Michael Sulmeyer reviews why cyber weapons have underperformed strategic expectations to date (unpredictable effects, complex target classification, scarce expert capacity) and assesses how agentic AI is poised to erode each constraint, unlocking previously theoretical strategic potential in cyber weapons.
Why it matters
CISOs and national security planners should recalibrate threat models: the paper's thesis implies offensive cyber capability may soon scale independent of human expert bottlenecks, changing assumptions underlying current cyber-defense investment and doctrine.
Action needed
Incorporate the paper's tactical-to-strategic escalation thesis into the next cyber threat model review and red-team scoping exercise.