In partnership with City University of Seattle. Apply Microsoft's STRIDE threat model to autonomous, tool-calling AI agents — six threat categories, six real attack scenarios, and the concrete mitigations to build in. An extension of the STRIDE Analysis on Agentic Systems talk.
Module 1
Spoofing
Module 2
Tampering
Module 3
Repudiation
Module 4
Information Disclosure
Module 5
Denial of Service
Module 6
Elevation of Privilege