Photo: IBM Technology / YouTube
OWASP's updated Top 10 for large language models reveals how easily AI systems can be manipulated, poisoned, or tricked into leaking sensitive data.
Claude Sonnet 4.6 blurs the line between mid-tier and flagship AI. What happens when capabilities outpace our ability to measure them?
Agent Zero's latest update lets anyone teach AI agents new tricks in minutes. The demo is impressive. The security warnings? Even more so.
IBM's Grant Miller explains how AI agents with elevated permissions create security nightmares—and what actually works to prevent privilege escalation.
Google's Model Armor adds security checkpoints to AI agents through ADK callbacks, intercepting threats before they reach language models.
The fastest-growing open source AI project reveals why agents that actually do things are both irresistible and architecturally dangerous.