AgentMisalignment: Measuring the Propensity for Misaligned Behaviour in LLM-Based Agents

Jason Ross Brown, Edward James Young, Akshat Naik, Patrick Quinn, Guillermo Bosch, Emma Gouné, Francisco Javier Campos ZabalaSubmission to ICLR 2026 · Rejected

Abstract

As Large Language Model (LLM) agents become more widespread, associated misalignment risks increase. While prior research has studied agents’ ability to produce harmful outputs or follow malicious instructions, it remains unclear how likely agents are to spontaneously pursue unintended goals in realistic deploy- ments. In this work, we approach misalignment as a conflict between the internal goals pursued by the model and the goals intended by its deployer. We introduce a misalignment propensity benchmark, AGENTMISALIGNMENT, a benchmark suite designed to evaluate the propensity of LLM agents to misalign in realistic scenarios. Evaluations cover behaviors such as goal-guarding, resisting shutdown, sandbagging, and power-seeking. Testing frontier models, we find that more capa- ble agents tend to exhibit higher misalignment on average. We also systematically vary agent personalities through different system prompts and observe that persona characteristics can strongly and unpredictably influence misalignment, sometimes more than the choice of model itself. Our results reveal the limitations of current alignment methods for autonomous LLM agents and underscore the need to rethink misalignment in realistic deployment settings.
BibTeX
@inproceedings{brown2026agentmisalignment,
  author = {Jason Ross Brown and Edward James Young and Akshat Naik and Patrick Quinn and Guillermo Bosch and Emma Gouné and Francisco Javier Campos Zabala},
  title = {{AgentMisalignment: Measuring the Propensity for Misaligned Behaviour in LLM-Based Agents}},
  booktitle = {AIWILD @ ICLR},
  year = {2026},
  url = {https://openreview.net/forum?id=X0ZWkYymsy},
}
Figure