Human Visitors to Wikipedia Drop by 8%: AI Abstracts and Social Platforms Alter How We Seek Info

发布时间：2025-10-20 19:53:44 | 责任编辑：吴昊 | 浏览量：5 次

Marshall Miller from the Wikimedia Foundation disclosed in his latest blog post that the human page views on Wikipedia have decreased by 8%. This data was discovered after updating the robot detection system, which identified that the unusually high traffic in May and June was mainly from robots specifically designed to bypass detection.
Miller attributed the drop in traffic to "the impact of generative AI and social media on how people access information." He pointed out two major trends: search engines are increasingly using generative AI to provide answers directly to users instead of linking to websites like Wikipedia; and younger generations tend to seek information on social video platforms rather than the open web. However, Google previously disputed the claim that "AI summaries reduce search traffic."
Miller stated that the Wikimedia Foundation welcomes "new ways for people to access knowledge" and believes that the encyclopedia's importance is not diminished as long as knowledge from Wikipedia continues to spread, even if visitors do not directly access the website. Wikipedia even tried to launch its own AI summary feature, but it was paused after complaints from editors.
However, this shift does pose risks, especially when people become less clear about the true sources of information. Miller noted, "As fewer people visit Wikipedia, there may be fewer volunteers to enrich and expand the content, and fewer individual donors supporting the work." He added that some volunteers have made outstanding contributions; reportedly, a volunteer subdued a gunman at a Wikipedia editing meeting last Friday.
For this reason, Miller advocates that AI, search, and social media companies using Wikipedia content should "encourage more visitors" to access the website itself.
Wikipedia is taking its own measures. The organization is developing a new content attribution framework to label content from the encyclopedia. In addition, it has formed two teams to help Wikipedia reach new readers and is recruiting volunteers to assist with this work.
Miller also called on readers to "support content integrity and content creation more broadly." He wrote, "When you search for information online, look for citation sources and click through to the original material. Discuss with people you know the importance of credible, manually curated knowledge, and help them understand that the content behind generative AI is created by real people who deserve support."
From an industry impact perspective, the decline in Wikipedia traffic reflects a fundamental change in content distribution models in the AI era. When search engines and AI assistants provide answers directly, users no longer need to visit the original content sources. This "mediation" trend presents structural challenges for open knowledge platforms that rely on traffic and community engagement.
The dilemma faced by Wikipedia is representative. As the largest collaborative knowledge base on the internet, its content is widely used to train AI models and generate search answers, but this usage severs the traffic feedback loop — users who don't visit the site won't become editors or donors, and in the long run, may threaten the sustainability of content production.
This has also sparked discussions about content attribution and value distribution. When AI systems use open resources like Wikipedia to provide services and gain commercial benefits, how to ensure that original content creators and platforms receive appropriate recognition and support becomes an urgent issue. Miller's content attribution framework and the suggestion to click on original sources represent an initial response from content providers to this issue.
From a user behavior perspective, there is a contradiction between the convenience of information access and the ability to trace information sources. AI summaries and social media feeds offer immediate answers, but they also reduce users' awareness of the sources, reliability, and completeness of information. This trend may have a negative impact on the long-term health of the online information ecosystem.
Welcome to the [AI Daily] column! This is your daily guide to exploring the world of artificial intelligence. Every day, we present you with hot topics in the AI field, focusing on developers, helping you understand technical trends, and learning about innovative AI product applications.