AI-Generated Content: Increasing the Burden on Wikipedia Editors

AI-Generated Content: Increasing the Burden on Wikipedia Editors

Blog Article

As AI-generated content increasingly permeates the user-generated Internet, driven by large language models (LLMs) like OpenAI’s GPT, it's important to consider the challenges faced by Wikipedia editors. Beyond their typical role of correcting inaccurate human edits, they now find themselves dedicating more time to identifying and removing AI-generated filler content.

404 Media spoke with Ilyas Lebleu, a Wikipedia editor and co-founder of the “WikiProject AI Cleanup” initiative. This group is working to establish best practices for detecting machine-generated contributions—despite the fact that AI tools are not effective in this area.

One significant issue with AI-generated content is its frequent lack of proper sourcing. The ability of LLMs to quickly generate large volumes of seemingly credible text has even resulted in entire fabricated entries being submitted, attempting to deceive Wikipedia's human experts.