Newer, Larger, Better? A Critique of the Unreflective LLM Adoption in Communication Research

The growing adoption of large language models (LLMs) in political communication research has prompted excitement but also concern. In this opinion piece, we offer an informed and critical overview of common LLM use cases in the field, including text analysis, synthetic data generation, and experiments. We argue that while these tools can be appealing, they often introduce serious epistemic, environ-mental, and infrastructural trade-offs that are insufficiently acknowledged. Beyond technical limitations, we highlight deeper issues related to scholarly autonomy, methodological opacity, resources inequality, and corporate dependency. Rather than dismissing innovation, we advocate for critical reflexivity and a renewed commitment to methodological rigor. While examining shortcomings of LLMs in current practices, we also point to viable alternatives. In essence, we call for a more deliberate, context-sensitive integration of LLMs in social science – one that prioritizes transparency, sustainability, and scientific integrity.