We introduce document retrieval and comment generation tasks for automating horizon scanning. This is an important task in the field of futurology that collects sufficient information for predicting drastic societal changes in the mid- or long-term future. The steps used are: 1) retrieving news articles that imply drastic changes, and 2) writing subjective comments on each article for others’ ease of understanding. As a first step in automating these tasks, we create a dataset that contains 2,266 manually collected news articles with comments written by experts. We analyze the collected documents and comments regarding characteristic words, the distance to general articles, and contents in the comments. Furthermore, we compare several methods for automating horizon scanning. Our experiments show that 1) manually collected articles are different from general articles regarding the words used and semantic distances, 2) the contents in the comment can be classified into several categories, and 3) a supervised model trained on our dataset achieves a better performance. The contributions are: 1) we propose document retrieval and comment generation tasks for horizon scanning, 2) create and analyze a new dataset, and 3) report the performance of several models and show that comment generation tasks are challenging.