Topic: production agents
-
11% of AI Agents Meet New Security Standards
Only 11% of 100 tested production AI agents met emerging security benchmarks, exposing critical vulnerabilities like credential exposure and privilege escalation. Many agents operate with excessive permissions and standing credentials, creating severe risks by allowing continuous access to sensit...
Read More » -
AI Agent Risk Assessment and Categorization Guide
The enterprise AI landscape is evolving from basic chatbots to autonomous "AI agents" capable of reasoning and taking actions, creating a new and urgent security paradigm focused on categorizing risk based on "access and autonomy". AI agents are categorized into three classes with distinct se...
Read More »