Wikipedia has inked major AI firms, including Microsoft, Meta and Amazon, to high-priority data access deals. The Wikimedia Foundation's commercial subsidiary, Wikimedia Enterprise, offers faster and more voluminous API access to its vast 65 million articles for a fee.
The agreements bring the total number of major tech companies that have partnered with the foundation's paid program to six, following Google in 2022. Smaller firms such as Ecosia, Nomic, Pleias, ProRata and Reef Media also participate.
Lane Becker, president of Wikimedia Enterprise, explained that Wikipedia content is "a critical component" for big tech companies' work, but needs financial support. The foundation has been exploring how to monetize its data without compromising its free ethos.
The push for paid access comes as infrastructure costs have risen sharply due to the increasing demand from AI developers who scrap Wikipedia content in large volumes. In April last year, bandwidth used for multimedia downloads grew by 50 per cent since January 2024, largely driven by bots. The foundation also reported a fall of around eight per cent in human traffic to its site over the past year.
In response, Wikipedia introduced new measures to detect and block automated scrapers, which has led to a decline in website visitors. However, founder Jimmy Wales welcomed AI models training on Wikipedia data, but said big tech companies should pay for their fair share of the costs incurred by maintaining the platform.
Wikipedia's revenue from its paid program will help offset infrastructure costs, which are largely funded by small donations. The deals demonstrate the growing importance of the foundation's content in the development of artificial intelligence models.
The agreements bring the total number of major tech companies that have partnered with the foundation's paid program to six, following Google in 2022. Smaller firms such as Ecosia, Nomic, Pleias, ProRata and Reef Media also participate.
Lane Becker, president of Wikimedia Enterprise, explained that Wikipedia content is "a critical component" for big tech companies' work, but needs financial support. The foundation has been exploring how to monetize its data without compromising its free ethos.
The push for paid access comes as infrastructure costs have risen sharply due to the increasing demand from AI developers who scrap Wikipedia content in large volumes. In April last year, bandwidth used for multimedia downloads grew by 50 per cent since January 2024, largely driven by bots. The foundation also reported a fall of around eight per cent in human traffic to its site over the past year.
In response, Wikipedia introduced new measures to detect and block automated scrapers, which has led to a decline in website visitors. However, founder Jimmy Wales welcomed AI models training on Wikipedia data, but said big tech companies should pay for their fair share of the costs incurred by maintaining the platform.
Wikipedia's revenue from its paid program will help offset infrastructure costs, which are largely funded by small donations. The deals demonstrate the growing importance of the foundation's content in the development of artificial intelligence models.
. On one hand, it's great that they're exploring ways to monetize their data without compromising their free ethos. But on the other hand, I'm worried about the impact on Wikipedia's community and traffic 
. With more companies using their content for AI development, we risk losing some of the platform's integrity and user base. It's a delicate balance between making money and preserving the core values that make Wikipedia so great 
.
they need to find a way to make ends meet but this feels like a trade-off? all these big tech firms are getting preferential access to their data, which is basically the backbone of wikipedia 
. doesn't that feel a bit unfair to the average user who just wants to learn stuff online for free? 
, it's time for them to step up and pay their share. i mean, wikipedia is all about the free knowledge, but it seems like no one wants to contribute financially 
. plus, its not like they're just taking the content willy-nilly, its being used for AI models and stuff which is actually pretty cool 
so i guess this is just one step in that direction 
seriously though, it's kinda cool that Wikipedia's finally finding ways to monetize its massive library without selling out 
. I mean, we all know AI dev's love scraping free content like it's going out of style 
, so it's nice to see the Wikimedia Foundation getting paid for their work... even if it means fewer humans visiting the site 
. At least Jimmy Wales is keeping it real about big tech companies taking advantage of the platform 
. Can't wait to see how this all plays out! 
. And now they're losin' visitors on purpose by blockin' automated scrapers? That's just a bummer 
... and if they can make it work with the big tech companies, who knows what kind of magic they'll come up with 
?