Corpus of public-facing communications by AI providers

This repository stores the textual dataset collected and used in the paper "Risk and Governance in Industry Communication around General-Purpose AI" (under review). The corpus is provided as csv files, each corresponding to a company, which contain the date of publication and link to each text. For copyright reasons, no text excerpts or full texts are included. To use the data, use the links to retrieve the texts. The corpus consists of blog posts and online articles (n = 4,110) from AI and GPU providers between January 2021–June 2024. If you use this data, please cite it as: Anonymous, 2026. Corpus of public-facing communications by AI providers. Zenodo. DOI: 10.5281/zenodo.17406089.