AI has become an ambiguous, yet powerful term that routinely refers to different technological advances, with stakeholders competing for public attention to influence public perceptions and decision-makers. While previous research has identified a dominant role of corporate actors and economic framings in discourses on AI, a systematic investigation of the role of a broad range of stakeholders within public discourses in the formative phase of current AI development is missing. Therefore, this paper maps out the AI stakeholder environment and offers a stakeholder typology for analysing this environment along two axes of stakeholder types and levels. This framework is used to investigate stakeholders in the AI Twitter discourse across the U.S. and Germany from 2012 until 2021. This early period of budding AI hype and political and economic framing arguably delineated country-specific trajectories that influence the current discourse and regulation of GenAI within these two nations. The results highlight the institutionalisation of stakeholders in the German discourse, with not only economic but also political organisations taking a strong role. The US dataset, in contrast, reveals strong shifts in stakeholder types and levels. Overall, the analysis offers insights into different cycles of institutionalisation in the early AI discourse on Twitter. It also highlights clear shifts in stakeholder involvement at different stages of public AI discourse in its formative phase in different socio-cultural and geographical contexts.

Corporate actors have been active in shaping the normative debates associated with the development and uptake of AI in the past years. We show how they achieve this by analysing corporate documents and public communication materials related to AI governance from 2017 to 2023 in China, Germany and the U.S. The analysis draws on the concept of sociotechnical imaginaries and shows that companies in the three countries have strategically constructed imaginaries such as “Responsible AI”, “Trustworthy AI” and “AI for Good” through communication, network-formation, and institution-building. These imaginaries both respond to regulatory pressure and shape public policy that serves companies’ interests. While these corporate imaginaries draw from local social and historical resources and respond to national policy agendas and public sentiments, they share the fact that they empower these corporations to dominate future AI governance infrastructure and institutionalisation. This finding calls for critical reflections on the corporate influence on the public negotiation of AI governance.

News media play a central role in the public negotiation of emerging technologies. This paper compares public debates and imaginaries of artificial intelligence in Chinese, German and U.S. news coverage between 2012 and 2024. Building on the concept of sociotechnical imaginaries and drawing from computational and qualitative analyses of articles in 15 leading national news media across China, Germany, and the U.S., the paper reveals that between the three countries, AI was most extensively discussed in Chinese coverage, where it was most prominently linked to health-, policy-, and technology topics. In contrast, German and U.S. media relate AI more often to its societal impacts. Sentiment toward these topics was more positive in the U.S., more negative in German, and comparatively neutral in Chinese AI coverage. Furthermore, the qualitative analysis demonstrates that even within topics, crucial differences in AI-related imaginaries can be found across countries. For example, in the healthcare context, AI is imagined to enhance healthcare in all three countries. Yet, while U.S. and German coverage envisioned enhancements on an individual patient level, Chinese newspapers described AI as a means to enhance public health service capacity at scale. In sum, this paper highlights the contextual nature as well as sociocultural and sociopolitical shaping of the public negotiation of sociotechnical futures with AI.

The discussant (anonymised) will discuss the panel’s presentations afterwards and discuss their broader implications. With his expertise in the analysis of public debates, comparative research and on sociotechnical innovations in the field of communication in general and AI in particular, the discussant is exceptionally well suited to provide the basis for a fruitful discussion.