We live in an era in which AI models succeed one another at a rapid pace, driven by commercial interests and market logic. The promise is great, but the question of who truly benefits — and who bears the costs — largely remains unanswered. It is precisely in this context that the concept of Public AI offers a valuable alternative.
Public AI starts from the conviction that AI systems can also be designed and governed in the public interest — similar to how we think about public broadcasting, public education, or public infrastructure. In doing so, it offers a way out of the apparent contradiction between uncritically going along with the hype on the one hand, and categorically rejecting AI on the other.
In this panel discussion, we explore what it concretely means to develop and deploy AI according to public values — so that knowledge of and access to culture, information, and democratic processes remain genuinely publicly accessible. Perspectives from academia, public broadcasting, the research community, and civil society are brought together. The central questions explored in the conversation are: How can we, across diverse sectors, formulate a shared understanding of Public AI, and what does that require of us in terms of investment, design choices, and collective action, both nationally and at a European level?