# Should I Trust You? Detecting Deception in Negotiations using Counterfactual RL

arXiv: https://arxiv.org/abs/2502.12436 · HTML: https://mlsanity.com/paper/2502.12436/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 69 references that are in the map.
- First posted on arXiv 18 Feb 2025.

## Builds on (most cited first, 23 of 23)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [The Internal State of an LLM Knows When It's Lying](https://mlsanity.com/paper/2304.13734/) (2023)
- [Release Strategies and the Social Impacts of Language Models](https://mlsanity.com/paper/1908.09203/) (2019)
- [A Survey of the State of Explainable AI for Natural Language Processing](https://mlsanity.com/paper/2010.00711/) (2020)
- [Language Models Learn to Mislead Humans via RLHF](https://mlsanity.com/paper/2409.12822/) (2024)
- [Deep Counterfactual Regret Minimization](https://mlsanity.com/paper/1811.00164/) (2018)
- ["Why is 'Chicago' deceptive?" Towards Building Model-Driven Tutorials for Humans](https://mlsanity.com/paper/2001.05871/) (2020)
- [Mastering the Game of No-Press Diplomacy via Human-Regularized Reinforcement Learning and Planning](https://mlsanity.com/paper/2210.05492/) (2022)
- [Evaluating Superhuman Models with Consistency Checks](https://mlsanity.com/paper/2306.09983/) (2023)
- [Finding Friend and Foe in Multi-Agent Games](https://mlsanity.com/paper/1906.02330/) (2019)
- [Human-Level Performance in No-Press Diplomacy via Equilibrium Search](https://mlsanity.com/paper/2010.02923/) (2020)
- [No Press Diplomacy: Modeling Multi-Agent Gameplay](https://mlsanity.com/paper/1909.02128/) (2019)
- [No-Press Diplomacy from Scratch](https://mlsanity.com/paper/2110.02924/) (2021)
- [Learning to Play No-Press Diplomacy with Best Response Policy Iteration](https://mlsanity.com/paper/2006.04635/) (2020)
- [Learning to Discuss Strategically: A Case Study on One Night Ultimate Werewolf](https://mlsanity.com/paper/2405.19946/) (2024)
- [Linguistic Harbingers of Betrayal: A Case Study on an Online Strategy Game](https://mlsanity.com/paper/1506.04744/) (2015)
- [Learning Strategic Language Agents in the Werewolf Game with Iterative Latent Space Policy Optimization](https://mlsanity.com/paper/2502.04686/) (2025)
- [More Victories, Less Cooperation: Assessing Cicero's Diplomacy Play](https://mlsanity.com/paper/2406.04643/) (2024)
- [To Tell The Truth: Language of Deception and Language Models](https://mlsanity.com/paper/2311.07092/) (2023)
- [Personalized Help for Optimizing Low-Skilled Users' Strategy](https://mlsanity.com/paper/2411.09109/) (2024)

## Built on by (most cited first, 3 of 3)

- [Shall We Play a Game? Language Models for Open-ended Wargames](https://mlsanity.com/paper/2509.17192/) (2025)
- [Measuring Fine-Grained Negotiation Tactics of Humans and LLMs in Diplomacy](https://mlsanity.com/paper/2512.18292/) (2025)
- [ChartAttack: Testing the Vulnerability of LLMs to Malicious Prompting in Chart Generation](https://mlsanity.com/paper/2601.12983/) (2026)

## Abstract (arXiv)

An increasingly common socio-technical problem is people being taken in by offers that sound ``too good to be true'', where persuasion and trust shape decision-making. This paper investigates how abrai can help detect these deceptive scenarios. We analyze how humans strategically deceive each other in textitDiplomacy, a board game that requires both natural language communication and strategic reasoning. This requires extracting logical forms of proposed agreements in player communications and computing the relative rewards of the proposal using agents' value functions. Combined with text-based features, this can improve our deception detection. Our method detects human deception with a high precision when compared to a Large Language Model approach that flags many true messages as deceptive. Future human-abrai interaction tools can build on our methods for deception detection by triggering textitfriction to give users a chance of interrogating suspicious proposals.