Wikipedia's Battle Against AI-Generated Content: Editors Mobilize to Maintain Reliability

Wikipedia's Growing Challenge with AI-Generated Content

Wikipedia, one of the world's most trusted online resources, is facing a new challenge in the form of AI-generated content. With the rise of large language models like ChatGPT, the platform has seen an influx of articles and edits containing false information, phony citations, and poorly written content 1

1

. This phenomenon, dubbed "AI slop" by the Wikipedia community, has prompted a robust response from the site's volunteer editors.

The Community's Response: WikiProject AI Cleanup

In response to this growing issue, Wikipedia editors have mobilized to create WikiProject AI Cleanup. This initiative aims to tackle the "increasing problem of unsourced, poorly written AI-generated content" 1

1

. The project has developed guidelines to help editors identify AI-generated text, including:

1. Excessive use of em dashes
2. Overuse of certain conjunctions like "moreover"
3. Promotional language such as "breathtaking"
4. Formatting issues like curly quotation marks instead of straight ones

A study by Princeton University researchers found that approximately 5% of new English-language Wikipedia pages created in August 2024 contained AI-generated text 2

2

. This highlights the scale of the problem and the need for vigilant monitoring.

New Policies: Speedy Deletion and AI Content Warnings

To combat the influx of AI-generated content, Wikipedia has implemented new policies. In early August, the platform amended its speedy deletion policy, allowing editors to quickly remove articles with obvious hallmarks of AI-generated writing 1

1

. This bypasses the typical seven-day discussion period for article removal, streamlining the process for clearly problematic content.

Additionally, Wikipedia has introduced a new warning label for articles suspected of incorporating AI-generated text. This label, which reads "This article may incorporate text from a large language model," has been affixed to hundreds of Wikipedia articles 2

2

.

The Role of the Wikimedia Foundation

While the Wikimedia Foundation, which hosts Wikipedia, doesn't create policies for the website, it is actively involved in developing tools to address the AI challenge. Marshall Miller, the Foundation's product director, likens the community's response to an "immune system" adapting to a new challenge 1

1

.

The Foundation is developing a non-AI-powered tool called Edit Check, designed to help new contributors adhere to Wikipedia's policies and writing guidelines. Features of Edit Check include:

1. Reminders to add citations for large amounts of text
2. Tone checks to ensure neutrality
3. A planned "Paste Check" feature to verify the originality of large text chunks 1

   1

Balancing AI Use and Human Oversight

Despite the challenges, the Wikimedia Foundation sees potential benefits in AI technology. They already use AI to help identify article revisions containing vandalism and are exploring ways to support editors with AI tools for automating repetitive tasks and translations 1

1

.

However, the Foundation emphasizes that human oversight remains crucial. "The way that Wikipedia remains neutral and reliable, and the thing that makes it unique, is that all this content passes through the hands of people," says Miller 2

2

.

The Ongoing Battle and Its Implications

The fight against AI-generated content on Wikipedia is ongoing, with editors working tirelessly to maintain the site's reliability. This struggle raises important questions about the role of AI in knowledge creation and dissemination. As Lucie-Aimée Kaffee, a policy and AI researcher, notes, "This will be a problem that will be in many and most aspects of our life. Asking, 'Where do we want AI? ... What kind of rules do we want to set for AI-generated knowledge?' I think is really important" 2

2

.

As Wikipedia continues to adapt to this new challenge, its experiences may provide valuable insights for other organizations grappling with the impact of AI on content creation and curation.