What Exploit Research Taught Me About AI Agent Security
Exploit research teaches security professionals to examine what systems can actually do—not merely what designers intended. Tim Shelton applies that principle to AI agent architecture, exploring combined permissions, runtime enforcement, human-controlled autonomy, progressive authority, and the need to verify operational outcomes.
Incident Response Metrics: Measure Time to Verified Outcome
Alert counts and response times don't tell executives whether a cyber threat was actually contained. David Harris examines the incident response lifecycle, including evidence gathering, decision latency, authorized action, and time to verified containment—and explains what CEOs should measure after a significant security incident.

