← All posts
security
3 posts ai security
+5
Building AI Agent-Aware Systems
Most systems can't detect when an AI agent is acting. Here's how to build the token claims, policy surfaces, and audit trail that changes that.
Read more
ai security
+4
AI Agent Session Security: Prompt Injection and Dry Runs
Browser-use agents can be hijacked by the pages they visit. Dry runs don't replicate production. Here's how to build safer sessions.
Read more
ai security
+4
AI Agent Identity Assurance: The NIST IAL/AAL Crosswalk
NIST's IAL/AAL was built for humans. When AI agents act on your behalf, the assurance levels still apply - the mapping just changes.
Read more