Defensive Sufficiency in a Stackelberg Model of AI Security
A Stackelberg economic model evaluates when feedback-driven security testing and repairs provide sufficient protection for AI systems.
A Stackelberg economic model evaluates when feedback-driven security testing and repairs provide sufficient protection for AI systems.