Topic · 23 episodes across 11 reviews

Agents on the Attack: Offensive Capability and Adversarial Robustness

← all reviews

Two security papers bookend the agent threat landscape — one where an autonomous agent finds and exploits 28 Windows zero-days, one where a sentence on a webpage traps agents in expensive infinite loops.

Covered in these reviews

Related concepts

Episodes and papers for this topic live on the concept pages.

AI Safety Agentic AI Prompt Injection AI & Security Agentic Misalignment Evaluation & Benchmarks LLM-as-Judge Multi-Agent Systems Tool Use Agent Scaffolding Agentic Coding Emergent Behavior Reward Hacking Scalable Oversight