• Being a doctoral researcher at the Research Unit

Communicative Construction of Alternative AI Futures at Algorithm Watch and Mozilla Foundation

PhD project of Phillip Engelhardt

AI is a technology that is constantly re-emerging in terms of its areas of application and its transformational potential. Its meaning remains flexible and a large part of the discourse surrounding the future of AI is therefore focused on filling the still-underexploited semantic space with communicatively constructed meanings. In this process of sense-making, diverse economic, political, and technological interests compete for relevance.

Non-governmental (NGOs) and non-profit organizations (NPOs) play a distinct role in this context, as their constructions of meaning are characterized by explicitly advocating public interests. For example, leading commercial large language models (LLMs) developed by companies such as OpenAI, Anthropic, Google DeepMind, Meta, DeepSeek, XAI, and Mistral open up various concerns and potentials at the interface with the public. The (re-)production of false statements and deceptions, as well as unethical, anti-democratic, and discriminatory responses, pose a threat to social cohesion and democratic deliberation. At the same time, LLMs offer new opportunities for political participation, scientific education, and the optimization of public administration.

The dissertation investigates which AI futures are envisioned by NGOs and NPOs, how these futures are developed, and what role the public plays in them. The organizations Algorithm Watch and Mozilla Foundation serve as case studies for the project. The dissertation particularly focuses on comparing futures that are developed and communicated as alternatives to commercial AI futures and demand the regulation of privatized AI or promote the development of AI as public property. The methodological basis of the project is a qualitative analysis of narrative expert interviews and communication materials (websites, newsletters, position papers, policy recommendations) according to Grounded Theory. Theoretically, the work understands alternative AI futures as the result of a process of communicative construction.

Contact

• Phillip Engelhardt