Wikimedia has called on AI companies to take responsibility for using Wikipedia content in their language models. This can be achieved by stopping scraping and using the paid API instead.
In a blog post, the organization states that artificial intelligence cannot exist without the human knowledge collected and maintained on platforms such as Wikipedia. To maintain that balance, Wikimedia asks developers of generative AI to clearly cite their sources and contribute to the continued existence of the open knowledge project via the paid Wikimedia Enterprise platform.
Wikimedia Enterprise enables companies to use Wikipedia data at scale without overloading servers. At the same time, they financially support the organization’s non-profit mission. Wikimedia emphasizes that it is not considering legal action against companies that scrape information from the site. However, the organization expects AI companies to continue using the content responsibly and sustainably.
Fewer human visitors
The call comes at a time when Wikipedia is facing changing traffic patterns. According to TechCrunch, the number of human visitors fell by eight percent compared to last year. The news site also reports that Wikimedia discovered that a large share of the traffic peak in May and June came from AI bots attempting to impersonate human users. After the organization adjusted its detection systems, it turned out that these bots were using Wikipedia on a large scale for training purposes.
In the blog post, Wikimedia states that transparency about the origin of information is essential for trust in online knowledge. When platforms do not indicate where their information comes from, both the quality and continuity of that knowledge are at risk. Less traffic to Wikipedia ultimately means fewer volunteers contributing and fewer donations to keep the project running.
The foundation also reiterates that it itself uses artificial intelligence to support volunteers, for example, in detecting vandalism or translating articles. This use is intended to make people’s work easier, not to replace them.
With this message, Wikimedia aims to make clear to the AI industry that the future of reliable technology cannot be separated from human knowledge production. Without continuous human contributions, the foundation on which AI is built will erode.