Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods

• URL: http://arxiv.org/abs/2505.05541v1
• Date: Thu, 08 May 2025 16:55:07 GMT
• Title: Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods
• Authors: Markov Grey, Charbel-Raphaël Segerie,
• Abstract summary: This literature review consolidates the rapidly evolving field of AI safety evaluations.<n>It proposes a systematic taxonomy around three dimensions: what properties we measure, how we measure them, and how these measurements integrate into frameworks.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: As frontier AI systems advance toward transformative capabilities, we need a parallel transformation in how we measure and evaluate these systems to ensure safety and inform governance. While benchmarks have been the primary method for estimating model capabilities, they often fail to establish true upper bounds or predict deployment behavior. This literature review consolidates the rapidly evolving field of AI safety evaluations, proposing a systematic taxonomy around three dimensions: what properties we measure, how we measure them, and how these measurements integrate into frameworks. We show how evaluations go beyond benchmarks by measuring what models can do when pushed to the limit (capabilities), the behavioral tendencies exhibited by default (propensities), and whether our safety measures remain effective even when faced with subversive adversarial AI (control). These properties are measured through behavioral techniques like scaffolding, red teaming and supervised fine-tuning, alongside internal techniques such as representation analysis and mechanistic interpretability. We provide deeper explanations of some safety-critical capabilities like cybersecurity exploitation, deception, autonomous replication, and situational awareness, alongside concerning propensities like power-seeking and scheming. The review explores how these evaluation methods integrate into governance frameworks to translate results into concrete development decisions. We also highlight challenges to safety evaluations - proving absence of capabilities, potential model sandbagging, and incentives for "safetywashing" - while identifying promising research directions. By synthesizing scattered resources, this literature review aims to provide a central reference point for understanding AI safety evaluations.

Related papers

• Probing and Steering Evaluation Awareness of Language Models [0.0]
  Language models can distinguish between testing and deployment phases.<n>This has significant safety and policy implications.<n>We show that linear probes can separate real-world evaluation and deployment prompts.
  arXiv  Detail & Related papers  (2025-07-02T15:12:43Z)
• Systematic Hazard Analysis for Frontier AI using STPA [0.0]
  frontier AI companies currently do not describe in detail any structured approach to identifying and analysing hazards.<n>A (Systems-Theoretic Process Analysis) is a systematic methodology for identifying how complex systems can become unsafe, leading to hazards.<n>We evaluateA's ability to broaden the scope, improve traceability and strengthen the robustness of safety assurance for frontier AI systems.
  arXiv  Detail & Related papers  (2025-06-02T15:28:34Z)
• Advancing Embodied Agent Security: From Safety Benchmarks to Input Moderation [52.83870601473094]
  Embodied agents exhibit immense potential across a multitude of domains.<n>Existing research predominantly concentrates on the security of general large language models.<n>This paper introduces a novel input moderation framework, meticulously designed to safeguard embodied agents.
  arXiv  Detail & Related papers  (2025-04-22T08:34:35Z)
• Toward an Evaluation Science for Generative AI Systems [22.733049816407114]
  We advocate for maturing an evaluation science for generative AI systems.<n>In particular, we present three key lessons: Evaluation metrics must be applicable to real-world performance, metrics must be iteratively refined, and evaluation institutions and norms must be established.
  arXiv  Detail & Related papers  (2025-03-07T11:23:48Z)
• AILuminate: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons [62.374792825813394]
  This paper introduces AILuminate v1.0, the first comprehensive industry-standard benchmark for assessing AI-product risk and reliability.<n>The benchmark evaluates an AI system's resistance to prompts designed to elicit dangerous, illegal, or undesirable behavior in 12 hazard categories.
  arXiv  Detail & Related papers  (2025-02-19T05:58:52Z)
• Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress? [59.96471873997733]
  We propose an empirical foundation for developing more meaningful safety metrics and define AI safety in a machine learning research context.<n>We aim to provide a more rigorous framework for AI safety research, advancing the science of safety evaluations and clarifying the path towards measurable progress.
  arXiv  Detail & Related papers  (2024-07-31T17:59:24Z)
• Evaluating General-Purpose AI with Psychometrics [43.85432514910491]
  We discuss the need for a comprehensive and accurate evaluation of general-purpose AI systems such as large language models. Current evaluation methodology, mostly based on benchmarks of specific tasks, falls short of adequately assessing these versatile AI systems. To tackle these challenges, we suggest transitioning from task-oriented evaluation to construct-oriented evaluation.
  arXiv  Detail & Related papers  (2023-10-25T05:38:38Z)
• Sociotechnical Safety Evaluation of Generative AI Systems [13.546708226350963]
  Generative AI systems produce a range of risks. To ensure the safety of generative AI systems, these risks must be evaluated. We propose a three-layered framework that takes a structured, sociotechnical approach to evaluating these risks.
  arXiv  Detail & Related papers  (2023-10-18T14:13:58Z)
• Evaluating Model-free Reinforcement Learning toward Safety-critical Tasks [70.76757529955577]
  This paper revisits prior work in this scope from the perspective of state-wise safe RL. We propose Unrolling Safety Layer (USL), a joint method that combines safety optimization and safety projection. To facilitate further research in this area, we reproduce related algorithms in a unified pipeline and incorporate them into SafeRL-Kit.
  arXiv  Detail & Related papers  (2022-12-12T06:30:17Z)
• Evaluating the Safety of Deep Reinforcement Learning Models using Semi-Formal Verification [81.32981236437395]
  We present a semi-formal verification approach for decision-making tasks based on interval analysis. Our method obtains comparable results over standard benchmarks with respect to formal verifiers. Our approach allows to efficiently evaluate safety properties for decision-making models in practical applications.
  arXiv  Detail & Related papers  (2020-10-19T11:18:06Z)
• SAMBA: Safe Model-Based & Active Reinforcement Learning [59.01424351231993]
  SAMBA is a framework for safe reinforcement learning that combines aspects from probabilistic modelling, information theory, and statistics. We evaluate our algorithm on a variety of safe dynamical system benchmarks involving both low and high-dimensional state representations. We provide intuition as to the effectiveness of the framework by a detailed analysis of our active metrics and safety constraints.
  arXiv  Detail & Related papers  (2020-06-12T10:40:46Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.