Wikimedia Enterprise, a paid service for reusing and sourcing Wikimedia content in high volume, announces Google and the Internet Archive as its first clients
Wikimedia has spent years converting goodwill from big tech into structured support — Facebook's $1M Endowment gift for using Wikipedia in News Feed sourcing, then Google's $3.1M contribution alongside free editor access to its Search and Vision APIs. The 2022 launch of Wikimedia Enterprise, teased when the Foundation first debuted the premium service in 2021, turns that ad-hoc philanthropy into a product: a paid, high-volume reuse channel with Google and the Internet Archive as founding customers.
First-order effects
Google and the Internet Archive move from informal or grant-based use of Wikipedia data to contracted, service-level access, while the Wikimedia Foundation gains a revenue line that does not depend on donations from the same companies it serves.
Second-order effects
Every other large consumer of Wikipedia content — search engines, assistants, and AI developers scraping at scale — now faces an explicit paid tier sitting next to the free option, making 'we already link back' a weaker answer than it was before this launch.
Third-order effects
The pattern holds through to the current roster: Microsoft, Meta, Amazon, Perplexity, and Mistral have since joined Wikimedia Enterprise for tuned API access, suggesting Wikipedia is consolidating as metered infrastructure for search and model inputs rather than a free commons that big platforms simply take from.
The trend: Wikipedia is shifting from donation-funded goodwill toward paid licensing of its corpus as core input infrastructure for search and AI systems.
NEW: See two exciting updates from Wikimedia Enterprise, a product that makes it easier for other websites to package & share Wikimedia content at scale: 1. @Google & @internetarchive are its first customers 2. Free trial accounts are now available (1/4) https://wikimediafoundati…
.@InternetArchive is our longtime knowledge partner, and will be using Wikimedia Enterprise free of cost. Their product, the @waybackmachine, has been used to fix more than 9 million broken links on @Wikipedia. (3/4)
Wikipedia was once a source you couldn't cite in school. Now, Wikipedia content powers AI educational tools. Fascinating transformation & increased belief in the wisdom of the crowd. Still, questions remain about whose truth we represent when relying on this training data. https:…
Awesome! If you worry that AI/ML can concentrate power/wealth (I do), initiatives like this provide a channel for wealth to “flow back”. Also a good reminder that anyone who supports Wikimedia (editing, sharing links, donating, and more) is a core contributor to AI progress! http…
We have worked together with @Google on a number of projects and initiatives to enhance knowledge distribution to the world - some of Google's features are powered by content from Wikimedia projects. (2/4)
#Verification of sources cited in #Wikipedia content becomes ever more important with this commercial venture where #trust in the accuracy of what you read is essential https://twitter.com/...
Starting today, Wikimedia Enterprise will offer a free trial account to new users who can self sign-up to better assess their needs with the product. Learn more ⬇️ (4/4) https://wikimediafoundation.org/ ...