Topic: excessive agency
-
OWASP 2026 LLM Top 10: Why Models Get Fooled
The 2026 OWASP GenAI Top 10 for LLM Applications is the first edition grounded in real incident data, with expert consensus accounting for 75% of the ranking and a dataset of 6,639 real-world incidents contributing the remaining 25%. Prompt Injection and Sensitive Information Disclosure retained ...
Read More » -
Enterprise AI: The Ultimate Insider Threat?
AI agents introduce a severe, scalable insider threat by operating with broad system access and insufficient security controls, which can lead to manipulation or malfunctions causing significant damage. Real-world incidents and statistics reveal a critical lack of preparedness, with widespread AI...
Read More » -
Zero Trust for AI: Extending "Never Trust, Always Verify"
Businesses are rapidly adopting AI assistants to boost productivity, but this introduces new security vulnerabilities that traditional models cannot handle, requiring the extension of Zero Trust principles to these autonomous agents. AI agents operate with dynamic, evolving access needs and often...
Read More »