Hype vs. Doom – How Can AI Reporting Become More Nuanced?

Artificial intelligence (AI) has become a cross-cutting topic that oscillates in public debate between technological promises of salvation and dystopian visions of the future. It is precisely this tension that makes reporting on AI so challenging. How do you report on a topic that is both ubiquitous and difficult to grasp? A Research Clinic at the Leibniz Institute for Media Research | Hans-Bredow-Institut (HBI), led by Mercator Fellow and technology journalist Svea Eckert, explored the challenges of AI reporting—and what needs to change to make it more nuanced. 

By Milena Braun and Svea Eckert 

The Research Clinic is an interactive, research-oriented format. It brought together journalists from a range of media organizations specializing in AI and technology to collaboratively examine how—and under what conditions—reporting on AI is produced in German newsrooms. The format created a space for reflection where different journalistic perspectives—from reporters and section editors to and editors-in-chief—could come together. 

The first discussion was based on concrete examples and experiences from editorial practice that illustrated both the conditions under which AI reporting is produced and the challenges currently facing it. The discussion was complemented by a research presentation by Dominic Lammar from the Technical University of Munich. He summarized how AI is portrayed in the German media and how these portrayals simultaneously shape expectations of AI in the present and the future. His presentation served as a mirror for journalistic practice and as a starting point for reflecting on editorial routines. Participants then developed future visions for AI reporting. The aim was to use a utopian framework to move beyond existing resource constraints and structural limitations. 

AI Reporting at the Intersection of Journalistic Organization 

The first part of the Research Clinic made it clear that the challenges of AI reporting begin with the basic question of how the topic is framed. Several participants emphasized that AI should no longer be understood as an isolated technological phenomenon, but rather as a cross-cutting societal issue that affects a wide range of domains. Nevertheless, many newsrooms continue to treat AI primarily as a standalone topic, typically assigning it to technology or digital desks. This editorial classification has a direct impact on reporting. In practice, in-depth AI coverage often depends on the specialization of individual journalists, as subject-matter expertise is usually developed through personal areas of focus and independent professional development. At the same time, participants explained that external perspectives from academia and professional practice have so far only been integrated to a limited extent. Where such resources are lacking, journalistic assessments of AI often remain superficial. 

Traditional News Values and Limited AI Expertise Hinder Critical Contextualization 

Another point of discussion was that AI reporting continues to be strongly shaped by traditional news values. Coverage is produced primarily when new, unexpected, or problematic developments emerge—for example, when publicly debated failures occur or when AI is introduced into areas of public administration. In day-to-day newsroom practice, this event-driven focus leaves little room for longer-term analysis and critical contextualization.  

Participants also noted that it is difficult to communicate the technical complexity of AI in journalistic reporting. This sometimes results in oversimplified and highly generalized portrayals. Against the backdrop of the limited specialist expertise already described, AI is therefore often used as an umbrella term encompassing a wide range of different technologies and applications. However, this conceptual vagueness makes nuanced reporting more difficult and encourages narratives that oscillate between technological promises of salvation and dystopian visions of the future. 

Generalized Reporting and Algorithms Fuel the AI Hype 

In this context, Dominic Lammar’s research presentation highlighted that the AI hype makes certain visions of the future appear particularly plausible, while other perspectives receive less public attention. Participating journalists observed that this dynamic is driven less by journalistic news values alone than by algorithmic logics of attention and distribution. They argued that the media have a responsibility to contextualize the different forms, manifestations, and power structures surrounding AI and make them understandable to a broader audience.  

This leads to another key challenge: communicating with audiences that possess widely differing levels of knowledge and experience. While AI has already become part of the professional or personal lives of some people, it remains abstract and difficult to grasp for others. Several participants expressed concern that parts of the audience had already been left behind—or risked being left behind in the future. 

Drawing on similar experiences from climate reporting, participants discussed how fundamental concepts are often assumed to be common knowledge, even though this is by no means the case. Good AI reporting must therefore do more than simply monitor technological developments or provide practical guidance. It must explain broader contexts, take different levels of knowledge into account, and highlight the societal implications of AI. The crucial question is not only what AI is technically capable of doing, but above all where and how it is already shaping society. 

Utopias for AI Reporting 

The second part of the Research Clinic focused on an intentionally open-ended question about the future: What would journalism—and AI reporting in particular—look like if resources were unlimited? The utopian scenarios developed by four groups served to identify concrete areas for future development and action.  

One key outcome was the shared vision of integrating AI as a cross-cutting topic throughout newsroom structures. Under this approach, AI would no longer be confined to rigid editorial beats but would instead be systematically incorporated into all areas of reporting. Closely connected to this was the idea of building multidisciplinary teams that would bring together journalistic, technical, and subject-specific expertise in a closer and more dynamic way. In addition to journalists and data reporters, participants specifically identified data scientists, software developers, video specialists, and subject-matter experts as integral members of editorial teams. Closer collaboration among these different groups of actors was seen as highly desirable. Participants also emphasized that greater transparency of algorithmic systems and data infrastructures would be a key prerequisite for more critical and reflective reporting. 

AI Reporting Should Also Be Enjoyable 

The utopian scenarios also addressed the structures of journalism itself. Participants repeatedly expressed a desire for more time and organizational resources for professional development—for example, through AI coaching—as well as for professional exchange and in-depth background discussions. At present, they noted, most continuing education and deeper engagement with the topic take place during journalists’  free time. 

 Closely linked to this was the vision of a workplace culture shaped less by publication pressure and more by opportunities for experimentation, stronger professional networks, and a more developed culture of learning from mistakes. Participants also discussed how the targeted use and further development of AI-supported tools could help journalists identify topics and make research processes more efficient. 

 Finally, questions of audience engagement and storytelling came into focus. Participants emphasized the importance of building closer connections with diverse audiences and of experimenting more boldly with new ways of communicating AI-related topics. The central idea was a form of reporting that not only explains complex technological developments but also translates them into different everyday contexts, thereby making them accessible to a wide range of audiences. AI reporting can, should, and is allowed to be enjoyable. 

Conclusion 

The Research Clinic demonstrated that the conditions for nuanced AI reporting do not depend solely on individual editorial decisions but are fundamentally shaped by the structural conditions under which journalism operates. Closer collaboration across editorial desks, additional resources for developing expertise, and stronger networks extending beyond individual newsrooms and publishing houses are therefore essential. It became clear that these conditions directly influence how AI is represented in the public sphere and which interpretive frameworks come to dominate. Through nuanced reporting on AI, journalism can help foster a better understanding of both technological and societal transformation while illuminating the space between narratives of hype and doom. 

Photos: © Leibniz Institute for Media Research / Milena Braun