Regulatory gray areas of LLM Terms

TL;DR

This paper analyzes the Terms of Service of five major LLM providers (Anthropic, OpenAI, Google, xAI, DeepSeek) as of November 2025 to reveal how usage restrictions vary and create regulatory gray areas affecting research. Using document analysis and cross-referencing, the authors compile a Table of Terms and identify implications for general users and for researchers in security, computational social science, and psychology, including high-stakes use and restrictive language around personal data and deception. The study highlights substantial variation in specificity, emphasizes the risk of a shifting liability model where enforcement relies on users, and advocates a standardized Research Use addendum to preserve independent research access. It also discusses policy implications, enforcement challenges, and suggests pathways for harmonization to support responsible AI research while safeguarding users. The work provides a publicly accessible cross-platform resource (OSF) to guide researchers and institutions.

Abstract

Large Language Models (LLMs) are increasingly integrated into academic research pipelines; however, the Terms of Service governing their use remain under-examined. We present a comparative analysis of the Terms of Service of five major LLM providers (Anthropic, DeepSeek, Google, OpenAI, and xAI) collected in November 2025. Our analysis reveals substantial variation in the stringency and specificity of usage restrictions for general users and researchers. We identify specific complexities for researchers in security research, computational social sciences, and psychological studies. We identify `regulatory gray areas' where Terms of Service create uncertainty for legitimate use. We contribute a publicly available resource comparing terms across platforms (OSF) and discuss implications for general users and researchers navigating this evolving landscape.