How Open Must Language Models be to Enable Reliable Scientific Inference?

2026-03-27

AI Summary

This study addresses the critical limitation posed by the insufficient openness of closed-source language models, which severely undermines the reliability of scientific inference. By systematically examining how the absence of model construction and deployment details compromises the validity of scientific reasoning, the work integrates scientific inference analysis, transparency assessment, and methodological scrutiny to propose a normative framework for guiding model selection and risk disclosure in research. The investigation reveals that most current closed-source models fail to meet the rigor required for credible scientific inquiry. The authors advocate for explicit justification of model choices and transparent articulation of strategies to mitigate inference-related risks, thereby offering a methodological foundation to enhance the trustworthiness of AI-driven scientific research.

Abstract

How does the extent to which a model is open or closed impact the scientific inferences that can be drawn from research that involves it? In this paper, we analyze how restrictions on information about model construction and deployment threaten reliable inference. We argue that current closed models are generally ill-suited for scientific purposes, with some notable exceptions, and discuss ways in which the issues they present to reliable inference can be resolved or mitigated. We recommend that when models are used in research, potential threats to inference should be systematically identified along with the steps taken to mitigate them, and that specific justifications for model selection should be provided.

Problem

• openness
• language models
• scientific inference
• model transparency
• reliable inference

Innovation

• openness
• scientific inference
• language models
• model transparency
• reliability