In the weeks around March 8, 2020, Spanish political authorities moved from denying and minimizing COVID-19 (veiling international recommendations) to establishing a State of Alarm. This uncertainty scenario is a natural experiment for exploring how concealment and diffusion of critical messages in official discourse affected public and published media, information transmission, and collective risk assessment. This study explores, through Natural Language Processing (NLP) and network theory, press, and Twitter agendas those days when (after international warnings, chaos on data, and the authorization of large demonstrations) Spain made the “alarming discovery” of COVID- 19. Results show a swift change in the climate of opinion, from the week before to the week after Women’s Day (March 8). Noninformation influenced agendas in terms of themes, feelings, and behaviors. The way different societies made COVID-19’s “discovery” became essential in the framing of the crisis and on the subsequent trust in authorities during the pandemic. The suppression of information in the first moments remains a key study question.