Wikipedia urges AI companies to use its paid API, and stop scraping

Wikipedia on Monday laid out a simple plan to ensure its website remains supported in the age of artificial intelligence, despite declining traffic.

In a blog post, the Wikimedia Foundation, the organization that runs the popular online encyclopedia, called on AI developers to use its content “responsibly” by ensuring their contributions are properly attributed and content is accessed through its paid product. Wikimedia platform for enterprise.

Subscribe, Paid product It allows companies to widely use Wikipedia content without “heavily taxing Wikipedia servers,” the Wikimedia Foundation blog says. The post explains. Additionally, the paid nature of the product allows AI companies to support the nonprofit’s mission.

While the post stops short of threatening penalties or any type of legal action for using material through scraping, Wikipedia Recently noticed That AI bots were scraping their website while trying to pass themselves off as human. After updating its bot detection systems, the organization found that unusually high traffic in May and June came from AI bots that were trying to “evade detection.” Meanwhile, it said “human page views” were down 8% year over year.

Now, Wikipedia is laying out its guidelines for AI developers and providers, saying that generative AI developers must provide attribution to give credit to the human contributors whose content it uses to create its output.

“For people to trust information shared online, platforms must make it clear where the information comes from and increase the chances of visiting and participating in those sources,” the post said. “With fewer visits to Wikipedia, fewer volunteers may be growing and enriching the content, and fewer individual donors may be supporting this work.”

Earlier this year, the organization issued AI strategy for editorswhich said it would use artificial intelligence to help editors workflow around tedious tasks, translation automation, and other tools that help its editors, not replace them.