This paper describes the SemEval 2018 shared task on semantic extraction from cybersecurity reports, which is introduced for the first time as a shared task on SemEval. This task comprises four SubTasks done incrementally to predict the characteristics of a specific malware using cybersecurity reports. To the best of our knowledge, we introduce the world's largest publicly available dataset of annotated malware reports in this task. This task received in total 18 submissions from 9 participating teams.