Step 1: Recall what an inverse response means.
An inverse response is a phenomenon in which the process output initially moves in a direction opposite to where it eventually settles, before reversing and moving toward its true final steady state value. This is a purely dynamic (transient) effect, not a stability issue.
Step 2: Analyze option A - positive zero.
Consider G(s) = K(1 - tau_a s) / [(tau_1 s + 1)(tau_2 s + 1)], with tau_a, tau_1, tau_2 > 0. This zero at s = 1/tau_a is a positive (RHP) zero. Final value theorem gives positive final value K. Initial slope is negative, meaning the response first dips before rising. This is the standard textbook cause of inverse response. Statement A is TRUE.
Step 3: Analyze option B - negative zero.
A negative (LHP) zero only modifies speed and overshoot, never causes a sign reversal. Statement B is FALSE.
Step 4: Analyze option C.
This is exactly the defining symptom of inverse response: y(t) moves one way initially, then reverses toward the opposite final steady state. Statement C is TRUE.
Step 5: Analyze option D.
Stability depends only on pole locations, not zeros. A RHP zero changes transient shape but never makes the process unstable. Statement D is FALSE.
Step 6: Conclusion.
A and C are necessarily true. \[ \boxed{A, C} \]