TL;DR:

  • Wikimedia Foundation asks AI companies to stop free scraping and pay for Enterprise API access
  • Wikipedia costs £179m annually to operate, funded primarily through donations
  • AI changing research habits threatens donation revenue as users bypass Wikipedia

The Wikimedia Foundation is calling on AI companies to cease free data scraping and begin paying for its Enterprise API, highlighting concerns about sustainability as artificial intelligence reshapes how people access information.

The Value Proposition

In a blog post Monday, the nonprofit organisation emphasised that AI companies require high-quality human-curated information to maintain model performance. Wikipedia’s extensive volunteer editor network ensures well-sourced content across over 300 languages—a resource the foundation argues merits compensation.

The seventh-most visited website globally, Wikipedia cost $179 million (approximately £142 million) to operate during the 2023-2024 fiscal year. The platform remains advert-free, relying primarily on donations to sustain operations.

Changing Research Behaviours

Wikimedia expressed concern that shifting research habits could undermine its funding model. As users increasingly turn to AI assistants like ChatGPT rather than Wikipedia for information, they bypass donation prompts displayed on the encyclopaedia’s homepage.

The Enterprise API would enable AI companies to “use Wikipedia content at scale and sustainably without severely taxing Wikipedia’s servers, whilst also enabling them to support our nonprofit mission,” according to the foundation.

Industry Context

The request comes amid broader tensions between content creators and AI firms. Publishers including Penske, The New York Times, and News Corp are pursuing copyright infringement lawsuits against AI companies. Conversely, organisations such as the Associated Press and Reuters have signed licensing agreements with AI firms.

Google previously agreed to a commercial access deal with Wikimedia in 2022. Representatives for other major AI companies—including OpenAI, Meta, Perplexity, Anthropic, Microsoft, DeepSeek, and xAI—did not immediately respond to requests for comment.

Source: CNET

Share this article