What happens when the people tasked with explaining agricultural science misunderstand it?  This completed classroom study examined how emerging science communicators formed judgments about gene-edited livestock during a deliberative learning activity, and how misinformation (including AI-amplified misinformation) shaped their reasoning.

We conducted a qualitative content analysis of transcripts from a 75-minute Town Hall simulation and accompanying written reflections in an undergraduate agricultural communication course (N = 30). In the simulation, students role-played stakeholders (e.g., producer, regulator, scientist, social-media influencer) and debated the use of gene-edited livestock in food and agricultural systems, with particular attention to its potential benefits, risks, regulatory implications, and societal impacts in a moderated town hall format. We analyzed transcripts and post-debate reflections to identify recurring factual inaccuracies.

Four recurring misconceptions about gene-edited livestock emerged from the analysis. These included: (1) reliance on outdated estimates of the annual economic burden of porcine reproductive and respiratory syndrome (PRRS); (2) inversion of standard definitions by describing gene editing as the insertion of foreign DNA; (3) circulation of an unverified “recall” narrative attributed to CRISPR off-target effects; and (4) misidentification of CRISPR Therapeutics as a livestock biotechnology firm. Subsequent analysis examined the sources through which these inaccuracies entered, most commonly through students’ use of generative AI tools during preparation, and circulated during discussion, as well as the ways they shaped students’ evaluations of gene-edited livestock, including perceptions of risk, urgency, and regulatory trust.

These findings suggest that the central instructional challenge is not simply teaching more facts but strengthening the epistemic habits that govern how future communicators decide what consider as reliable knowledge. This work offers an evidence-based model for integrating active learning and AI literacy to improve students’ capacity to communicate contested agricultural science accurately and credibly.