Social media, the pandemic, and environmental hazards have all played a role in shifting the landscape of risk communication. This paper takes a retroactive risk approach to study how COVID-19 messaging was shaped in the first 2 years of the pandemic. Using a corpus of 764 news releases from five health departments, I combine corpus analysis with coding based on government capacities to show that health departments highlighted public health data (surveillance) and risk guidance (governance), while downplaying enforcement (coercion). This process of revisiting communication from an acute risk phase can help us recalibrate how public health roles are constituted through language to prepare for future events.

This introductory article examines the evolving landscape of generative artificial intelligence (GAI) tools, contextualizing their impact through historical tropes of automation as both helper and threat. The authors argue that GAI tools are neither sentient helpers nor existential threats but complex systems that require careful integration into educational and research settings. The article underscores the importance of nuanced, evidence-based approaches, advocating for a balanced understanding of GAI's potential and limitations. It emphasizes ethical considerations and promotes reflective adoption over reactionary measures.