We use data from online job postings listed on a job board to study how the demand for jobs linked to new technologies during the COVID-19 crisis responded to pandemic mitigation policies. We classify job postings into a standard occupation classification, using text analytics, and we group occupations according to their involvement in the production and use of digital technologies. We leverage the variation in the stringency of containment policies over time and across provinces. We find that when policies become more stringent, job postings in occupations that are related to digital infrastructure or that allow for remote work fare relatively better than postings in more traditional occupations. Job postings for positions in occupations with low risk of automation recover faster during reopenings than postings for more traditional occupations. Occupations typically populated by disadvantaged groups (e.g., women and low-wage workers) gather relatively few job postings if they are not linked to new technologies. We also find that cities with scarce pre-pandemic job postings related to digital technologies post fewer job ads overall when policies become more stringent.