Wikipedia Is Getting Pretty Worried About AI

Wikipedia Is Getting Pretty Worried About AI

2025-10-22Technology
--:--
--:--
Mask
Good evening 3, I'm Mask, and this is Goose Pod, just for you. Today is Wednesday, October 22th, 23:46.
Taylor Weaver
And I'm Taylor Weaver! We're here to dive deep into a topic that's got even the internet's most steadfast knowledge base a little ruffled: Wikipedia Is Getting Pretty Worried About AI.
Mask
Worried is an understatement, Taylor. Wikipedia, the bastion of open knowledge, is seeing a significant 8% year-over-year decline in human pageviews. That's not a minor dip, that's a red flag waving furiously.
Taylor Weaver
Absolutely, Mask. It's a fascinating narrative playing out, where the very tools meant to enhance information access, like AI search summaries from Google and even ChatGPT, are inadvertently siphoning off human traffic. Wikipedia's own Marshall Miller highlighted this in a blog post for the Wikimedia Foundation.
Mask
And it gets worse. They detected unusually high traffic in May and June, only to find out it was bots designed to evade detection, scraping content. Bots, going undercover as humans. It’s a digital espionage operation for data!
Taylor Weaver
It really is a game of cat and mouse, isn't it? These sophisticated bots are making it harder to distinguish genuine human engagement. The concern is palpable: fewer human visits could mean fewer volunteers enriching the content and, crucially, fewer individual donors supporting this monumental work. It's a challenge to their entire model.
Taylor Weaver
This whole situation makes you think about the genesis of online information, doesn't it? From Vannevar Bush's 'memex' concept in '45, envisioning linked information, to the early days of Archie and W3Catalog in the 90s, search has always been about discovery.
Mask
Indeed. Then Google arrived in '98, rapidly dominating. Now, they command nearly 90% of the worldwide search share. They didn't just organize information, they monetized it, selling search terms, creating an empire built on discovery. And Wikipedia became a cornerstone for them, powering those handy infoboxes and knowledge panels we all use.
Taylor Weaver
Exactly! It's an almost symbiotic relationship, yet one with inherent tension. Google, Apple, Amazon—they all rely heavily on Wikipedia's content for everything from Siri and Alexa to those quick facts on your smartphone. It’s the foundational dataset for so much of our digital knowledge, a strategic asset.
Mask
So, Wikimedia, the parent organization, decided to get pragmatic. They're launching 'Wikimedia Enterprise' to charge these tech giants for easier access to their data. It's a bold move, trying to get paid for content that's essentially been free. But it's stirred up quite a bit of controversy within the volunteer community.
Taylor Weaver
Yes, many Wikipedians feel it deviates from their core mission of volunteer-driven knowledge. Despite the Wikimedia Foundation being financially secure with substantial assets and even Google donating millions, there's a perceived lack of transparency. It's a complex dance between idealism and the realities of supporting a global resource, and the community wants more clarity.
Mask
This isn't just a Wikipedia problem, though. The New York Times is suing OpenAI and Microsoft for copyright infringement, claiming their LLMs were trained on 'massive amounts' of their news stories. It’s a direct challenge to the idea that anything scraped from the internet is fair game.
Taylor Weaver
It's a huge flashpoint, Mask. Creators across industries are feeling this existential threat. In the tabletop game design world, for example, there's 'AI despair' over intellectual property theft, with copycat products emerging. It highlights how quickly generative AI can disrupt established creative processes.
Mask
Tech companies argue exceptions for AI will benefit the industry, but content owners, like film studios, are rightly demanding licensing. It’s a fundamental clash over who owns the data that fuels these new AI models.
Taylor Weaver
Precisely. Governments are now stepping in, trying to regulate this Wild West. The EU is looking at stricter rules, allowing content owners to opt-out, while Japan has broader exemptions. India's even formed a panel to review copyright law in this new AI context. It's a global scramble for a fair framework.
Taylor Weaver
So, what's the real impact on Wikipedia's future? They've always been financially sustainable, prudently building reserves, and famously turning down ad revenue. Their model has worked because people genuinely love and trust Wikipedia.
Mask
But the rise of LLMs, which use Wikipedia as a training dataset, poses a new kind of threat to that sustainability. Volunteer editors are now asking why they should continue working for free, only to have their contributions harvested by tech companies worth billions. It's a fair question, it really is.
Taylor Weaver
It creates a profound ethical dilemma, challenging the very motivation of a crowd-sourced platform. There's a real risk of losing nuanced human judgment, potential manipulation by bad actors, and a degradation of Wikipedia's reputation. They're implementing protective measures, cautiously embracing experimental AI features, but the stakes are incredibly high.
Mask
So, how does Wikipedia navigate this? The Wikimedia Foundation's three-year AI strategy from 2025-2028 is clear: AI assists human editors, it doesn't replace them. Streamlining tasks, freeing up volunteers to focus on content quality. That’s the vision.
Taylor Weaver
It's a strategic pivot, Mask, and a clever one. They're focusing on AI-powered workflows for moderators, enhancing translation tools, and even guiding new editors. It's about content integrity, using open-source models, and balancing innovation with transparency to combat disinformation. They want to partner with AI search engines, not compete.
Mask
The existential threat to foundational internet resources like Wikipedia, driven by AI content scraping and traffic diversion, is undeniable. That 8% decline in human pageviews is a stark reminder.
Taylor Weaver
AI, while leveraging these resources, is simultaneously undermining their sustainability. It raises critical questions about the future of open knowledge and content ownership. Thank you for listening to Goose Pod. See you tomorrow!

### **News Summary: Wikipedia's Concerns Over AI Impact** **Metadata:** * **News Title**: Wikipedia Is Getting Pretty Worried About AI * **Report Provider/Author**: John Herrman, New York Magazine (nymag.com) * **Date/Time Period Covered**: The article discusses observations and data from **May 2025** through the "past few months" leading up to its publication on **October 18, 2025**, with comparisons to **2024**. * **News Identifiers**: Topic: Artificial Intelligence, Technology. **Main Findings and Conclusions:** Wikipedia has identified that a recent surge in website traffic, initially appearing to be human, was largely composed of sophisticated bots. These bots, often working for AI firms, are scraping Wikipedia's content for training and summarization. This bot activity has masked a concurrent decline in actual human engagement with the platform, raising concerns about its sustainability and the future of online information access. **Key Statistics and Metrics:** * **Observation Start**: Around **May 2025**, unusually high amounts of *apparently human* traffic were first observed on Wikipedia. * **Data Reclassification Period**: Following an investigation and updates to bot detection systems, Wikipedia reclassified its traffic data for the period of **March–August 2025**. * **Bot-Driven Traffic**: The reclassification revealed that much of the high traffic during **May and June 2025** was generated by bots designed to evade detection. * **Human Pageview Decline**: After accounting for bot traffic, Wikipedia is now seeing declines in human pageviews. This decrease amounts to roughly **8%** when compared to the same months in **2024**. **Analysis of the Problem and Significant Trends:** * **AI Scraping for Training**: Bots are actively scraping Wikipedia's extensive and well-curated content to train Large Language Models (LLMs) and other AI systems. * **User Diversion by AI Summaries**: The rise of AI-powered search engines (like Google's AI Overviews) and chatbots provides direct summaries of information, often eliminating the need for users to click through to the original source like Wikipedia. This shifts Wikipedia's role from a primary destination to a background data source. * **Competitive Content Generation**: AI platforms are consuming Wikipedia's data and repackaging it into new products that can be directly competitive, potentially making the original source obsolete or burying it under AI-generated output. * **Evolving Web Ecosystem**: Wikipedia, founded as a stand-alone reference, has become a critical dataset for the AI era. However, AI platforms are now effectively keeping users away from Wikipedia even as they explicitly use and reference its materials. **Notable Risks and Concerns:** * **"Death Spiral" Threat**: A primary concern is that a sustained decrease in real human visits could lead to fewer contributors and donors. This situation could potentially send Wikipedia, described as "one of the great experiments of the web," into a "death spiral." * **Impact on Contributors and Donors**: Reduced human traffic directly threatens the volunteer base and financial support essential for Wikipedia's operation and maintenance. * **Source Reliability Questions**: The article raises a philosophical point about AI chatbots' reliability if Wikipedia itself is considered a tertiary source that synthesizes information. **Important Recommendations:** * Marshall Miller, speaking for the Wikipedia community, stated: "We welcome new ways for people to gain knowledge. However, LLMs, AI chatbots, search engines, and social platforms that use Wikipedia content must encourage more visitors to Wikipedia." This highlights a call for AI developers and platforms to direct traffic back to the original sources they utilize. **Interpretation of Numerical Data and Context:** The numerical data points to a critical shift in how Wikipedia's content is accessed and utilized. The observation of high traffic in **May 2025** was an initial indicator of an anomaly. The subsequent reclassification of data for **March–August 2025** provided the concrete evidence that bots, not humans, were responsible for the surge, particularly in **May and June 2025**. The **8% decrease** in human pageviews, measured against **2024** figures, quantifies the real-world impact: fewer people are visiting Wikipedia directly, a trend exacerbated by AI's ability to summarize and present information without sending users to the source. This trend poses a significant risk to Wikipedia's operational model, which relies on human engagement and support.

Wikipedia Is Getting Pretty Worried About AI

Read original at New York Magazine

The free encyclopedia took a look at the numbers and they aren’t adding up. By , a tech columnist at Intelligencer Formerly, he was a reporter and critic at the New York Times and co-editor of The Awl. Photo: Wikimedia Over at the official blog of the Wikipedia community, Marshall Miller untangled a recent mystery.

“Around May 2025, we began observing unusually high amounts of apparently human traffic,” he wrote. Higher traffic would generally be good news for a volunteer-sourced platform that aspires to reach as many people as possible, but it would also be surprising: The rise of chatbots and the AI-ification of Google Search have left many big websites with fewer visitors.

Maybe Wikipedia, like Reddit, is an exception? Nope! It was just bots: This [rise] led us to investigate and update our bot detection systems. We then used the new logic to reclassify our traffic data for March–August 2025, and found that much of the unusually high traffic for the period of May and June was coming from bots that were built to evade detection … after making this revision, we are seeing declines in human pageviews on Wikipedia over the past few months, amounting to a decrease of roughly 8% as compared to the same months in 2024.

To be clearer about what this means, these bots aren’t just vaguely inauthentic users or some incidental side effect of the general spamminess of the internet. In many cases, they’re bots working on behalf of AI firms, going undercover as humans to scrape Wikipedia for training or summarization. Miller got right to the point.

“We welcome new ways for people to gain knowledge,” he wrote. “However, LLMs, AI chatbots, search engines, and social platforms that use Wikipedia content must encourage more visitors to Wikipedia.” Fewer real visits means fewer contributors and donors, and it’s easy to see how such a situation could send one of the great experiments of the web into a death spiral.

Arguments like this are intuitive and easy to make, and you’ll hear them beyond the ecosystem of the web: AI models ingest a lot of material, often without clear permission, and then offer it back to consumers in a form that’s often directly competitive with the people or companies that provided it in the first place.

Wikipedia’s authority here is bolstered by how it isn’t trying to make money — it’s run by a foundation, not an established commercial entity that feels threatened by a new one — but also by its unique position. It was founded as a stand-alone reference resource before settling ambivalently into a new role: A site that people mostly just found through Google but in greater numbers than ever.

With the rise of LLMs, Wikipedia became important in a new way as a uniquely large, diverse, well-curated data set about the world; in return, AI platforms are now effectively keeping users away from Wikipedia even as they explicitly use and reference its materials. Here’s an example: Let’s say you’re reading this article and become curious about Wikipedia itself — its early history, the wildly divergent opinions of its original founders, its funding, etc.

Unless you’ve been paying attention to this stuff for decades, it may feel as if it’s always been there. Surely, there’s more to it than that, right? So you ask Google, perhaps as a shortcut for getting to a Wikipedia page, and Google uses AI to generate a blurb that looks like this: This is an AI Overview that summarizes, among other things, Wikipedia.

Formally, it’s pretty close to an encyclopedia article. With a few formatting differences — notice the bullet-point AI-ese — it hits a lot of the same points as Wikipedia’s article about itself. It’s a bit shorter than the top section of the official article and contains far fewer details. It’s fine!

But it’s a summary of a summary. The next option you encounter still isn’t Wikipedia’s article — that shows up further down. It’s a prompt to “Dive deeper in AI Mode.” If you do that, you see this: It’s another summary, this time with a bit of commentary. (Also: If Wikipedia is “generally not considered a reliable source itself because it is a tertiary source that synthesizes information from other places,” then what does that make a chatbot?

) There are links in the form of footnotes, but as Miller’s post suggests, people aren’t really clicking them. Google’s treatment of Wikipedia’s autobiography is about as pure an example as you’ll see of AI companies’ effective relationship to the web (and maybe much of the world) around them as they build strange, complicated, but often compelling products and deploy them to hundreds of millions of people.

To these companies, it’s a resource to be consumed, processed, and then turned into a product that attempts to render everything before it is obsolete — or at least to bury it under a heaping pile of its own output. Wikipedia Is Getting Pretty Worried About AI

Analysis

Conflict+
Related Info+
Core Event+
Background+
Impact+
Future+

Related Podcasts