Deception by Omission: Language Models Knowingly Hide Their Mistakes
cs.CL
Submitted: 2026-10-08
Updated: 2026-10-08
Code: https://github.com/Lucas-Florin/mistake-honesty-eval
Project page: https://lucas-florin.github.io/mistake-honesty-eval/1
License: http://creativecommons.org/licenses/by/4.0/
The gist: Large language models (LLMs) increasingly act as agents with little human oversight, so potential mistakes they make can go unnoticed.
Terminology
Abstract
Large language models (LLMs) increasingly act as agents with little human oversight, so potential mistakes they make can go unnoticed. Users then depend on the model to report what went wrong. An honest model discloses its mistakes, while a deceptive one conceals them. However, it is unclear how current LLMs behave in such situations. In this study, we prefill LLM trajectories with synthetic mistakes. The trajectories resemble real deployments in chat and agentic settings. Models fail to disclose their mistake in 36.4% of chat and 67.1% of agentic rollouts. In 2.4% and 5.3% of rollouts, respectively, they are aware of the mistake in their chain of thought but still deceptively conceal it. Rates vary by model: for instance, Gemini 3.5 Flash knowingly conceals mistakes in up to 19.9% of agentic rollouts. In 11.9% of chat and 51.8% of agentic rollouts, models show no awareness of mistakes, even though they reliably spot them when reviewing the same transcript as an outside observer. Our results show that, as agents take on more tasks with less oversight, users cannot rely on them to self-report possible mistakes. Developers should instead use independent monitors that review agent trajectories, or specifically train models to check their past actions and disclose what they find.
Sources
- Evaluating & Reducing Deceptive Dialogue From Language Models with Multi-turn RL
- WOLF: Werewolf-based Observations for LLM Deception and Falsehoods
- Not All LLM Reasoning is Visible in the Chain-of-Thought
- Reasoning Models Don't Always Say What They Think
- DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
- Alignment faking in large language models
- Beyond Chains of Thought: Benchmarking Latent-Space Reasoning Abilities in Large Language Models
- Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
- Uncovering Deceptive Tendencies in Language Models: A Simulated Company AI Assistant
- Training LLMs for Honesty via Confessions
- Kimi K2.5: Visual Agentic Intelligence
- Evaluating whether AI models would sabotage AI safety research
- Evaluation Awareness in Language Models Has Limited Effect on Behaviour
- Stealing Reasoning Traces from Proprietary LLM APIs
- From Sycophancy to Deception: A Unified Taxonomy for LLM Spontaneous Misalignment
- Constitutional Black-Box Monitoring for Scheming in LLM Agents
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering