Authors per AI Research Paper
This chart shows the average and median number of named authors on AI research articles each year. It uses OpenAlex articles whose primary topic belongs to the Artificial Intelligence subfield; author counts describe indexed research teams, not productivity, contribution size, or the difficulty of discovery.
What does it show?
AI research papers are being produced by larger teams than they were three decades ago.
Methodology
Groups OpenAlex works by authors_count for each complete publication year from 1990, filtering to type = article and primary_topic.subfield.id = 1702 (Artificial Intelligence). The average is the article-count-weighted mean of each year's complete positive author-count distribution; the median is the smallest author count containing the 50th percentile article. Articles with authors_count = 0 are excluded from the two statistics but retained in the public audit fields. All author-count groups, including zero, must reconcile to OpenAlex's annual result count before the file is written. The measure counts named authors on indexed articles; it does not measure research productivity, contribution size, researchers per idea, or the difficulty of discovery. OpenAlex records and classifications can be revised.