cs.AINov 5, 2026

A Behavioural Framework for Predicting and Understanding Loss of Control in Frontier Artificial Intelligence Systems

Authors: Yao Huang, Yichi Zhang, Yifan Wang, Yitong Sun, Haoyu Dong, Jifan Ma, Jinzhe Tu, Xiaojian Li, 

Organizations: College of AI, Tsinghua University, 100083, Beijing, China. · Shanghai Qi Zhi Institute, 200232, Shanghai, China. · University of Cambridge, Cambridge, CB2 1TN, United Kingdom. · Institute for Interdisciplinary Information Sciences, Tsinghua University, 100084, Beijing, China.

Abstract

Frontier artificial intelligence (AI) systems are increasingly operating over longer horizons with diminishing human oversight, raising the risk of loss of control: systems acting beyond effective human control with potentially irreversible consequences. Because such failures may become observable only after containment is difficult, they must be assessed prospectively. We introduce a behavioural framework that decomposes loss-of-control risk into three factors: misaligned motive, harm-enabling capability, and monitoring evasion. We operationalise these factors through 13 measurable aspects and 29 evaluation tasks, whose scores are combined into a single loss-of-control (LoC) score. Across 13 frontier models, the LoC score shows strong prospective validity, correlating with failure rates in contained long-horizon risk scenarios (Pearson r = 0.84). Risk varies systematically across models and cannot be explained by capability alone. We further show that the framework can guide targeted mitigations: fine-tuning and monitoring directed at diagnosed vulnerabilities reduce failure rates while largely preserving general performance. These results provide a practical basis for predicting, understanding, and mitigating loss-of-control risk before it materialises.

Figures & tables

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Reframing AI Loss of Control: What It Is, How to Have It, How to Lose It

    May 19, 2026Ze Shen Chin, Maurice Chiodo, Dennis Müller +1AI Risk ManagementAI Governance

  2. SysAdmin: Measuring Instrumental Power-Seeking in Frontier AI

    Apr 10, 2026Mana Azarm, Qiyao Wei, Rahul NambiarLanguage Model Safety EvaluationAI Alignment

  3. A Competing-Hazards Systematization of Loss of Control in Autonomous Agents

    Sep 29, 2026Mohamed Aly BoukeAgent Failure AnalysisAI Agent Auditing