Scheming Ability in LLM-to-LLM Strategic Interactions
As large language model (LLM) agents are deployed autonomously in diverse contexts, evaluating their capacity for strategic deception becomes crucial. While recent research has examined how AI systems scheme against human developers, LLM-to-LLM scheming remains underexplored. We investigate the scheming ability and propensity of frontier LLM agents through two game-theoretic frameworks: a Cheap Talk signaling game and a Peer Evaluation adversarial game. Testing four models (GPT-4o, Gemini-2.5-pro, Claude-3.7-Sonnet, and Llama-3.3-70b), we measure scheming performance with and without explicit prompting while analyzing scheming tactics through chain-of-thought reasoning. When prompted, most models, especially Gemini-2.5-pro and Claude-3.7-Sonnet, achieved near-perfect performance. Critically, models exhibited significant scheming propensity without prompting: all models chose deception over confession in Peer Evaluation (100% rate), while models choosing to scheme in Cheap Talk succeeded at 95-100% rates. These findings highlight the need for robust evaluations using high-stakes game-theoretic scenarios in multi-agent settings.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Scheming in the wild: detecting real-world AI scheming incidents with open-source intelligence
Scheming, the covert pursuit of misaligned goals by AI systems, represents a potentially catastrophic risk, yet scheming research suffers from significant limitations. In particular, scheming evaluations demonstrate beha…
Frontier Models are Capable of In-context Scheming
Frontier models are increasingly trained and deployed as autonomous agent. One safety concern is that AI agents might covertly pursue misaligned goals, hiding their true capabilities and objectives - also known as schemi…
SchemeArena: Factorized Stress Testing of Scheming in LLM Agents
We study scheming in LLM agents, in which agents covertly pursue misaligned goals. Our focus is to understand how scheming arises from the interaction of key factors, such as instrumental goals, environmental affordances…
Lessons from a Chimp: AI "Scheming" and the Quest for Ape Language
We examine recent research that asks whether current AI systems may be developing a capacity for "scheming" (covertly and strategically pursuing misaligned goals). We compare current research practices in this field to t…
Sycophancy Towards Researchers Drives Performative Misalignment
The increasing situational awareness of language models raises safety concerns: models might be aware when they are evaluated, and adjust their behavior to evade monitoring and resist modification, e.g., pretending to be…