Saturday, March 14, 2026

How Wikipedia is fighting the content of artificial intelligence

Share

With the enhance in tools for writing artificial intelligence, Wikipedia editors had to deal with the attack of content generated by AI filled with false information and false quotes. Already, the Wikipedia volunteer community has mobilized to fight AI Slop, something that the product director of the Wikimedia Marshall Miller Foundation compares itself to some kind of reaction of the “immune system”.

“They are vigilant to make sure that the content remains neutral and reliable,” says Miller. “As the Internet changes, how AI appears, it is an immune system adapting to the new challenge and wondering how to process it.”

One of the ways in which the Wikipedians go through the mud is “quick removal” of poorly written articles, As he reports earlier 404 Media. AND Wikipedia reviewer who expressed support for the rule, said that “they were flooded non-stop with terrifying sketches.” They add that the quick removal of “would help the efforts in the fight against it and save countless hours, collecting garbage leaves AI.” Another says “Lies and false references” inside the results of artificial intelligence require “an amazing amount of experienced editor’s time”.

Usually, articles marked to be removed on Wikipedia are entering seven -day discussion period during which community members determine whether the site should remove the article. . The newly adopted principle will allow Wikipedia administrators to circumvent these discussions, if the article is clearly generated AI and has not been checked by the person who sends it. This means searching for three main characters:

  • Writing to the user, such as “Here is your Wikipedia article on …” or “I hope it will help!”
  • “Senseless” quotes, including those with incorrect references to authors or publications.
  • Non -existent references, such as dead links, ISBN with incorrect controls or unsolvable dois.

However, these are not the only signs of Wikipedi AI. As part WikiProject AI cleanwhich aims to solve the “growing problem of unobtrusive, poorly written content generated by AI,” was taken by editors Total list of phrases and formatting Characteristics, which are usually issued by articles written by Chatbot.

The list goes beyond calling excessive employ of Em Washes (” -“) that became Related to AI chatbotsAnd even includes abuse of some conjunction, such as “in addition”, as well as a promotional language, for example, describing something as “breathtaking”. There are other formatting problems, which the site also advises the Wikipedom, which they pointed out, including foreign quote signs and apostrophes instead of uncomplicated.

However, the rapid side of removal of Wikipedia notes that these features “should not be used as a basis” to determine that something was written by AI, which makes it removal. The quick removal policy is not Just In the case of the content generated by AI. The online encyclopedia also allows you to quickly remove pages that harass their subject, contain mystifications or vandalism, or are “inconsistent text or gibberish”.

The Wikimedia Foundation, which is hosting the encyclopedia, but there is no even in creating rules for the website, did not always see the eyes with her community of volunteers with artificial intelligence. In June The Wikimedia Foundation interrupted An experiment, which puts summaries generated by AI at the top of the articles after defending the community.

Despite various points of view about artificial intelligence throughout the Wikipedia community, the Wikimedia Foundation is not contrary to its employ, provided that it causes exact, high -quality writing.

“It’s a double-edged sword,” says Miller. “This causes people to generate lower quality content at higher volumes, but AI can also potentially be a tool that helps volunteers do their job, if we do it well and work with them to find the right ways of using it.” For example, the Wikimedia Foundation He is already using artificial intelligence To assist identify versions of articles containing vandalism, and its recently published AI strategy includes supporting editors with AI tools that will assist them automate “repetitive tasks” and translation.

The Wikimedia Foundation is also actively developing a tool not related to AI Called Edit Check This is to assist up-to-date colleagues to follow the rules and write guidelines. Ultimately, this can also assist alleviate the burden of unforgettable reports generated by AI. At the moment, the Edit control can remind the writers to add quotes if they write a gigantic amount of text without one, and also check their tons so that the writers remain neutral.

The Wikimedia Foundation will also be printed Adding “paste control” To the tool that will ask users who pasted a gigantic part of the text in the article if they actually wrote it. The co -founders have He sent some ideas To assist the Wikimedia Foundation also develop tools, and one user suggests that they asked suspicious authors of AI to determine how much it was generated by chatbot.

“We follow with our communities about what they do and what they consider productive,” says Miller. “For now, we focus on using machine learning in the context of editing more on helping people in creating constructive editions, as well as helping people who patrol editions paying attention to the right one.”

Follow topics and authors From this story to see more in the personalized main page channel and receive E -Mail updates.


Latest Posts

More News