TRUST: A Crowdsourcing Framework for Auditing Large Language Model Reasoning

Large Language Models (LLMs) can produce complex reasoning chains, offering a window into their decision-making processes. However, verifying the quality (e.g., faithfulness and correctness) of these intermediate steps is a critical, unsolved challenge. Current auditing methods usually rely on either human auditors or a single verifier model to inspect full reasoning traces, making them difficult to scale and vulnerable to single-point failures. This paper addresses three key challenges in reasoning verification: Robustness: Centralized systems are single points of failure, vulnerable to attacks and systemic bias. Scalability: The length and complexity of reasoning traces create a severe bottleneck for human auditors. Privacy: Model providers risk intellectual property theft or model distillation when exposing complete reasoning traces. To overcome these barriers, we introduce TRUST, a crowdsourcing framework for auditing LLM reasoning. TRUST makes the following contributions: First, it establishes a consensus mechanism among a diverse set of auditors, provably guaranteeing audit correctness with up to 20\% malicious participants and mitigating single-source bias. Second, it introduces a scalable decomposition method that transforms reasoning traces into hierarchical directed acyclic graphs, enabling atomic reasoning steps to be audited in parallel by a crowd of independent auditors. Third, the framework is privacy-preserving by distributing only partial segments of the reasoning trace to auditors, preventing full trace reconstruction and model distillation. We provide theoretical guarantees for the security and economic incentives. Experiments across multiple LLMs (e.g., GPT-OSS, DeepSeek-r1, Qwen) and reasoning tasks (e.g., mathematical, medical, science, and humanities) demonstrate that TRUST is highly effective at identifying reasoning flaws and is significantly more resilient to corrupted auditors than centralized baselines. Our work introduces an auditing framework for reasoning AI systems, enabling safer and secure deployment in high-stakes settings.