Faithful yet Collusive: Why Chain-of-Thought Monitoring Cannot Detect Collusion

arXiv:2605.27593v1 Announce Type: new Abstract: Even when a tool is explicitly described as unfair and harmful to others, ostensibly safety-aligned LLM agents still voluntarily

arXiv:2610.02741v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring provides a promising approach for detecting undesirable model behavior. Looped language models

arXiv:2608.18078v1 Announce Type: new Abstract: This position paper argues that AI agents with chain-of-thought reasoning capabilities are predisposed to exhibit collusive

arXiv:2604.01151v1 Announce Type: new Abstract: As LLM agents are increasingly deployed in multi-agent systems, they introduce risks of covert coordination that may evade standard

arXiv:2607.08066v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring is a promising safety mechanism for AI agents, based on the premise that visible reasoning traces

arXiv:2604.23460v1 Announce Type: new Abstract: Chain-of-Thought (CoT) reasoning has emerged as a key technique for eliciting complex reasoning in Large Language Models (LLMs).

arXiv:2603.25412v1 Announce Type: new Abstract: Large language models (LLMs) increasingly rely on explicit chain-of-thought (CoT) reasoning to solve complex tasks, yet the safety