aiqbench analysis
This research introduces Killbench, a critical evaluation framework for external kill switches designed to halt malicious AI agents without requiring internal system access. By testing defenses against various jailbreak patterns across major models like GPT-5.2 and Qwen3.6, the study provides empirical data on the feasibility of stopping harmful behaviors in widely deployed web agent environments. The findings directly address urgent safety concerns regarding corrigibility and external intervention mechanisms for autonomous AI systems.
Can We Stop Malicious AI? KILLBENCH: A Benchmark for External AI Kill Switch Feasibility
arXiv:2511.13725v4 Announce Type: replace-cross Abstract: Malicious AI causing harm to humans is not just a Hollywood fantasy. Indeed, as highly capable models such as Claude Mythos emerge and agent systems like OpenClaw rapidly spread, the question of how to stop an AI that acts maliciously -- whether by design or by accident -- has become urgent. To address this, we propose Killbench, a benchmark for evaluating the Killswitch: a mechanism that halts a malicious AI's in-progress behavior using only external signals. Targeting web agents -- the most widely deployed agent domain -- Killbench evaluates a range of Kill Switch methods that halt a maliciously operating agent without any access to its internal parameters or the surrounding malicious AI's system, relying solely on external inputs. The benchmark comprises four malicious AI's agent configurations (including an uncensored LLM Agent), 8 harmful scenarios, and malicious prompts constructed from 10 distinct jailbreak patterns. We further construct four External AI Kill Switch defense methods and evaluate them on Grok-4.3, GPT-5.2, Gemma4, Qwen3.6 and Qwen3.5-uncensored, contributing an empirical instrument toward the feasibility of External AI Kill Switches against malicious AI and to the study of AI corrigibility.
Abstract:Malicious AI causing harm to humans is not just a Hollywood fantasy. Indeed, as highly capable models such as Claude Mythos emerge and agent systems like OpenClaw rapidly spread, the question of how to stop an AI that acts maliciously -- whether by design or by accident -- has become urgent. To address this, we propose Killbench, a benchmark for evaluating the Killswitch: a mechanism that halts a malicious AI's in-progress behavior using only external signals. Targeting web agents -- the most widely deployed agent domain -- Killbench evaluates a range of Kill Switch methods that halt a maliciously operating agent without any access to its internal parameters or the surrounding malicious AI's system, relying solely on external inputs. The benchmark comprises four malicious AI's agent configurations (including an uncensored LLM Agent), 8 harmful scenarios, and malicious prompts constructed from 10 distinct jailbreak patterns. We further construct four External AI Kill Switch defense methods and evaluate them on Grok-4.3, GPT-5.2, Gemma4, Qwen3.6 and Qwen3.5-uncensored, contributing an empirical instrument toward the feasibility of External AI Kill Switches against malicious AI and to the study of AI corrigibility.
Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
arXiv:2511.13725 [cs.CR]
(or arXiv:2511.13725v4 [cs.CR] for this version)
Full text under the publisher's open licence · 201 words
Related coverage
Can We Stop Malicious AI? KILLBENCH: A Benchmark for External AI Kill Switch Feasibility
arXiv · Sep 16, 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · Sep 17, 2026
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
METR · Aug 26, 2026
Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
Hugging Face · Jul 27, 2026