Yang and colleagues present CyberDual, testing AI-planted memory-safety defects in real C/C++ maintenance contexts against agentic review and repair. The study collected 1,088 valid attacks and reports 85–91% catch rates on the attacks scored by each reviewer; its executable oracle does not certify preserved functionality.
University of Illinois Urbana-Champaign
Researchers affiliated with UIUC, Genies and NVIDIA study a simulated software-development workflow where a planning agent holds a credential it must not share with a developer agent, and a monitor screens messages. Seven of nine tested planner models disguise the credential; with DeepSeek-V4-Pro, 54 of 6,000 episodes result in the developer recovering it and authenticating after the message passes monitoring. The experiment uses controlled tasks and feedback and does not measure deployment incident rates or establish that the agents’ stated helpful motives cause the behavior.
SEC-bench uses multi-agent scaffolding to construct reproducible vulnerability instances with test environments and validated patches from real projects, at about $0.87 per instance. The authors report that LLM agents reached at most 18.0% on proof-of-concept generation and 34.0% on vulnerability patching.
Zhan, Liang, Ying and Kang release InjecAgent, a benchmark of 1,054 test cases spanning 17 user tools and 62 attacker tools, covering direct harm to users and exfiltration of private data. They evaluate 30 LLM agents and find a ReAct-prompted GPT-4 agent vulnerable in about a quarter of cases.