Textual and Comparative Analyses on AI Policies: How Big Tech and Their Global Watchdogs Frame Virtues and Responsibility*

The rapid advancement of AI has raised pressing moral considerations, prompting many stakeholders to publish guidelines for responsible development. However, the specific virtues organizations prioritize in their AI policies, and the extent to which they acknowledge AI’s potential role as a moral agent, remain underexplored. We conducted a textual and comparative analysis of AI principles issued by Big Tech corporations (OpenAI, Google, Microsoft and Anthropic) and their global watchdogs. The analysis revealed: (1) commonalities and distinctions in the technomoral virtues emphasized across organizations, drawing on Vallor (2016)’s framework; (2) variation in how specifically those virtues are articulated in AI policies; and (3) divergent views on AI’s role to discuss responsibility–some framing AI purely as a tool, others hinting at AI’s moral status. These findings provide insights into the ethical frameworks guiding AI governance and highlight potential gaps in current AI policy discourse.