Currently, 25% of senior corporate security leaders say third-party risk management is largely run with agentic AI; this rises to 50% in two years.