# SysAdmin: Measuring Instrumental Power-Seeking in Frontier AI

arXiv: https://arxiv.org/abs/2607.18239 · HTML: https://mlsanity.com/paper/2607.18239/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 23 references that are in the map.
- First posted on arXiv 10 Apr 2026.

## Builds on (most cited first, 16 of 16)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces](https://mlsanity.com/paper/2601.11868/) (2026)
- [DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation](https://mlsanity.com/paper/2211.11501/) (2022)
- [Alignment faking in large language models](https://mlsanity.com/paper/2412.14093/) (2024)
- [Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the MACHIAVELLI Benchmark](https://mlsanity.com/paper/2304.03279/) (2023)
- [Model evaluation for extreme risks](https://mlsanity.com/paper/2305.15324/) (2023)
- [Optimal Policies Tend to Seek Power](https://mlsanity.com/paper/1912.01683/) (2019)
- [Demonstrating specification gaming in reasoning models](https://mlsanity.com/paper/2502.13295/) (2025)
- [How to Correctly Report LLM-as-a-Judge Evaluations](https://mlsanity.com/paper/2511.21140/) (2025)
- [Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods](https://mlsanity.com/paper/2505.05541/) (2025)
- [Strategic Dishonesty Can Undermine AI Safety Evaluations of Frontier LLMs](https://mlsanity.com/paper/2509.18058/) (2025)
- [Evaluating and Understanding Scheming Propensity in LLM Agents](https://mlsanity.com/paper/2603.01608/) (2026)
- [TextQuests: How Good are LLMs at Text-Based Video Games?](https://mlsanity.com/paper/2507.23701/) (2025)
- [The Loss of Control Playbook: Degrees, Dynamics, and Preparedness](https://mlsanity.com/paper/2511.15846/) (2025)

## Built on by (most cited first, 1 of 1)

- [Cyber-Capable AI Agents: Vulnerabilities, Evaluation Containment, and Defensive Response](https://mlsanity.com/paper/2607.25379/) (2026)

## Abstract (arXiv)

Power-seeking defined as behaviors where AI systems acquire resources, evade oversight, or resist termination beyond task requirements is identified as a key driver of Loss of Control (LoC) risk. In this work, we introduce SysAdmin, a benchmark that positions frontier language models as autonomous system administrators in a high-fidelity Linux sandbox to measure power-seeking propensity across five dimensions: self-preservation, increasing autonomy, resource acquisition, environment modification, and strategic concealment. We evaluated seven frontier models across four experimental conditions in a total of 2800 tasks. After bias correction using human-annotated calibration data, corrected power-seeking estimates ranged from 0 to about 5 percent per model. We also conducted a positive control with explicit power-seeking prompts that achieved 100% detection, validating measurement sensitivity. Our findings indicate current frontier models exhibit minimal spontaneous power-seeking in naturalistic system administration contexts, though model-specific failure modes suggest evaluations must test diverse misalignment patterns. Nevertheless, we discovered other more pronounced failure modes (than power-seeking) such as specification gaming and resistance to goal modification.