What is Safety? Corporate Discourse, Power, and the Politics of Generative AI Safety

TL;DR

The paper investigates how leading GenAI firms publicly frame safety, arguing that safety discourse functions as a sociotechnical and political practice that distributes responsibility while shaping governance expectations. Using critical discourse analysis on 75 corporate texts from OpenAI, Google, and Anthropic, the authors reveal patterns of authority construction, diffusion of accountability, and a dynamic, globally resonant risk frame. The study contributes to AI literacy and governance scholarship by proposing a discursive toolkit to help users critically evaluate corporate safety narratives and by highlighting how participatory rhetoric can mask symbolic governance. The findings underscore the need for critical scrutiny of corporate framings in HCI and policy work to advance accountability, equity, and justice in GenAI deployment.

Abstract

This work examines how leading generative artificial intelligence companies construct and communicate the concept of "safety" through public-facing documents. Drawing on critical discourse analysis, we analyze a corpus of corporate safety-related statements to explicate how authority, responsibility, and legitimacy are discursively established. These discursive strategies consolidate legitimacy for corporate actors, normalize safety as an experimental and anticipatory practice, and push a perceived participatory agenda toward safe technologies. We argue that uncritical uptake of these discourses risks reproducing corporate priorities and constraining alternative approaches to governance and design. The contribution of this work is twofold: first, to situate safety as a sociotechnical discourse that warrants critical examination; second, to caution human-computer interaction scholars against legitimizing corporate framings, instead foregrounding accountability, equity, and justice. By interrogating safety discourses as artifacts of power, this paper advances a critical agenda for human-computer interaction scholarship on artificial intelligence.