Understanding the Impact of AI Generated Content on Social Media: The Pixiv Case

TL;DR

This work investigates how AI Generated Content reshapes a large image-based social platform, using Pixiv's explicit tagging of AI versus human artwork to compare production and consumption dynamics. By assembling a public dataset of over 15 million artworks and 937 thousand creators (including 2.48 million AI-generated pieces), the study traces changes in activity, creator behavior, and content themes from 2022 to 2024, and examines per-content engagement. Key findings show a 50% rise in new artworks after AIGC introduction, a shift toward adult and female-themed content, faster AI-artist production but no major rise in total AI artwork output, and more uniform consumption for AI content with top human artworks retaining dominance in views and bookmarks. The results offer actionable implications for platform policy, community formation, and recommender system design, emphasizing the need to preserve diversity and support human creators while managing AI-driven content growth across online art communities.

Abstract

In the last two years, Artificial Intelligence Generated Content (AIGC) has received significant attention, leading to an anecdotal rise in the amount of AIGC being shared via social media platforms. The impact of AIGC and its implications are of key importance to social platforms, e.g., regarding the implementation of policies, community formation, and algorithmic design. Yet, to date, we know little about how the arrival of AIGC has impacted the social media ecosystem. To fill this gap, we present a comprehensive study of Pixiv, an online community for artists who wish to share and receive feedback on their illustrations. Pixiv hosts over 100 million artistic submissions and receives more than 1 billion page views per month (as of 2023). Importantly, it allows both human and AI generated content to be uploaded. Exploiting this, we perform the first analysis of the impact that AIGC has had on the social media ecosystem, through the lens of Pixiv. Based on a dataset of 15.2 million posts (including 2.4 million AI-generated images), we measure the impact of AIGC on the Pixiv community, as well as the differences between AIGC and human-generated content in terms of content creation and consumption patterns. Our results offer key insight to how AIGC is changing the dynamics of social media platforms like Pixiv.

Figures (10)

• Figure 1: Top to bottom: (a) Weekly count of new artworks; (b) Number of comments generated under the artworks uploaded on each week; (c) Weekly proportion of AI-generated artworks, and the proportion of views, bookmarks, and comments received by AI-generated artworks uploaded in the same week. (d) Weekly count of new creators; (e) Weekly count of active creators categorized by the age of the account; (f) Weekly count of artworks categorized by the age of the creator's account. (g) Proportion of restricted (adult) artworks; (h) Proportion of artworks with female-tags; (i) Proportion of artworks with male-tags.
• Figure 2: Monthly Gini index for tags on the number of artworks associated with the tag.
• Figure 3: (a) CDF of the difference between the upload times of two consecutive artworks by the same creator; (b) CDF of the number of artworks of each creator; (c) Distribution of artwork uploaded by creators; (d) Count of active days of a creator vs. the count of artworks created by the same creator.
• Figure 4: (a) Jobs of creators; (b) Percentage of artworks that include a link to an external platform.
• Figure 5: (a) Count of views and bookmarks in the 50th, 75th, 90th, and 99th percentiles in relation to the age of the artwork; (b) Boxplot of the views and bookmarks per artwork, the whiskers are 5th and 95th percentiles.