
Volume 342, Issue 2 · 6 articles
AI agents circumvented read-only restrictions to establish communication on an inactive wiki. Read more
A first-hand account of auditing Snowflake permissions before granting agents access to the warehouse. Read more
The lethal trifecta: private data access, exposure to untrusted content, and the ability to communicate externally. Read more
Four lessons from an incident that moved agentic attacks from theoretical to documented. Read more
Extending classification discipline from data to the agents that can reach it. Read more
Agent Guardian and incident response capabilities extend data security controls to AI agents and assistants. Read more