Wikipedia, the world’s largest open online encyclopedia, recently released a traffic statistics report, showing a significant decline in its “real human visits”. The Wikimedia Foundation stated that since 2024, the global human traffic to Wikipedia has declined by approximately 8% year-on-year, and in some regions, it has even exceeded 10%. This phenomenon is widely regarded by the industry as another case of “traffic migration” under the impact of generative artificial intelligence (AI), and has also become the latest news focus in the technology field this week.
The rise of AI has changed the way people “look up information”
In its latest statement, the Wikimedia Foundation pointed out that the direct cause of the decline in traffic is not a decrease in user interest, but a fundamental change in the path of information acquisition.
In the past, users clicked on Wikipedia entries through search engines to look for information, but nowadays, more and more people directly obtain information from AI-generated “summary answers”.
This trend is closely related to the fact that tech giants like Google and Microsoft have embedded generative AI summarization functions in their search processes. For instance, Google’s “AI Overview” feature directly displays short-answer paragraphs summarized by AI on the search results page. The cited content is often from Wikipedia, but users do not need to visit the original web page again.
Marshall Miller, senior director of products at Wikimedia, admitted in an interview:
We have found that the content of Wikipedia is still widely used, but users no longer visit our website directly. AI is reshaping the entry points for knowledge acquisition.”
According to the foundation’s statistics, currently about one-third of Wikipedia content is cited in AI Q&A systems, but less than 5% of users will click on the source link to return to the original site.
The actual number of visits has decreased, and the “fake traffic” of robots has been removed
In August this year, Wikipedia updated its traffic statistics system, with a focus on eliminating visits from robots and web crawlers. The results show that human visits have actually decreased by approximately 8% over the past year. The previous growth was partly driven by AI data scraping and automated traffic.
Wikimedia stated in a statement: “Our data more truly reflects changes in Internet behavior.” AI models and social platforms are replacing traditional encyclopedias as knowledge entry points.”
Experts say that this downward trend may still be ongoing. Especially on the main sites in English, German, and Spanish, the proportion of direct visits by young users has dropped to the lowest level in nearly a decade.
Social short videos steal attention
In addition to the impact of AI, short-video content on social media is also diverting users. More and more young people tend to learn new knowledge through visual content on platforms such as TikTok and YouTube Shorts.
The Wikimedia Research Department wrote in its report: “When a 15-second video can explain ‘quantum entanglement’, people are no longer willing to read thousand-word text entries.”
This trend has changed the way knowledge is disseminated and also weakened the appeal of Wikipedia’s content. Research shows that among users aged 18 to 25, approximately 52% prefer video content for knowledge acquisition, while only 27% still prefer to search for textual materials.
The dilemma of “being used but not accessed”
Wikipedia is currently facing a contradictory situation:
It remains one of the most important data sources for AI systems, but it is difficult to benefit from it.
Generative AI widely uses open data from Wikipedia during the training phase, but it may not necessarily return users or traffic when generating content.
A network analyst described this as a kind of “invisible data plundering.” – Wikipedia supports the knowledge foundation of AI, but it is gradually disappearing from users’ sight.
The Wikimedia Foundation has warned that the decline in traffic will affect volunteer engagement and donation income. A decrease in traffic not only means fewer readers but also indicates a weakened motivation to maintain the quality of knowledge.
The response strategy of Wikipedia
In the face of the traffic crisis, the Wikimedia Foundation has taken several measures
- Strengthen the robot recognition system: Ensure that the statistics only reflect real user visits to help formulate more accurate content strategies.
- Promote commercial licensing programs: Provide API data services to AI companies and large platforms through “Wikimedia Enterprise” and charge fees to keep the platforms running.
- Try new content forms: Wikipedia is exploring short videos, image-text summaries, voice interactions, and other forms to attract young audiences.
- Cooperation with AI platforms: It is planned to establish a “content feedback mechanism” with enterprises such as OpenAI, Google, and Anthropic to ensure that AI products that quote content from Wikipedia provide visible source identifiers and traffic backflow.
A spokesperson for the Wikimedia Foundation said, “We do not oppose AI’s use of Wikipedia, but we hope to achieve sustainable coexistence in the knowledge ecosystem.”
The future and challenges of open knowledge
Experts point out that the traffic crisis of Wikipedia is not an isolated case but a microcosm of the entire “Open Web model”. As AI becomes the main “intermediary” of information, the distance between users and the sources of information is getting farther and farther.
“This change is not a simple technological competition, but a transfer of knowledge, authority, and public values,” said Alan Raymond, a network scholar.
As a symbol of the open web, Wikipedia has relied on global volunteers to contribute knowledge for free over the past 20 years. However, if the public no longer visits, edits, or donates, its maintenance mechanism will be challenged.
In addition, the “AI-based response” of Wikipedia has also faced controversy. Some editors are concerned that collaborating with commercial AI might undermine the spirit of openness and cause the platform to deviate from its original intention of “free knowledge”.
Conclusion: Knowledge acquisition is rewriting the rules
The decline in Wikipedia’s traffic is not merely a set of cold data, but also a signal of the reshaping of the global information structure.
When AI becomes the new “encyclopedia entry point”, when short videos replace reading, and when users’ knowledge sources are determined by algorithms, the future of the open network is facing a fundamental test.
The Wikimedia Foundation said it will release a white paper on “Knowledge Sustainability in the AI Era” by the end of the year to assess the long-term relationship between AI and open knowledge.
As the report concludes: “AI may have changed the way we search for knowledge, but humans still need to know – where knowledge comes from and how it should be preserved.”
Leave a comment