Part 1 of 3: The “Great Acceleration of Risk” is here—and your trusted AI agents may be your biggest vulnerability