Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

highplainsdem

(61,398 posts)
Wed Mar 4, 2026, 11:57 AM 5 hrs ago

AI Translations Are Adding 'Hallucinations' to Wikipedia Articles

Source: 404 Media

Wikipedia editors have implemented new policies and restricted a number of contributors who were paid to use AI to translate existing Wikipedia articles into other languages after they discovered these AI translations added AI “hallucinations,” or errors, to the resulting article.

The new restrictions show how Wikipedia editors continue to fight the flood of generative AI across the internet from diminishing the reliability of the world’s largest repository of knowledge. The incident also reveals how even well-intentioned efforts to expand Wikipedia are prone to errors when they rely on generative AI, and how they’re remedied by Wikipedia’s open governance model.

The issue in this case starts with an organization called the Open Knowledge Association (OKA), a non-profit organization dedicated to improving Wikipedia and other open platforms.

-snip-

Wikipedia editors investigated how OKA was operating and found that it was mostly relying on cheap labor from contractors in the Global South, and that these contractors were instructed to copy/paste articles to popular LLMs to produce translations.

-snip-



Read more: https://www.404media.co/ai-translations-are-adding-hallucinations-to-wikipedia-articles/



No surprise that this went badly.

The article mentions a job posting from OKA offering $397/mo for working up to 40 hours a week, translating as many as 20 articles per week. OKA originally instructed its translators to use Grok, but has since changed that recommendation to other chatbots.

Any of which can also hallucinate at any time.
12 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies

Martin68

(27,490 posts)
1. I have great respect and appreciation or those who work to keep Wikipedia objective and truthful. It's difficult job in
Wed Mar 4, 2026, 12:10 PM
5 hrs ago

this environment, and there are many who would love to subvert and corrupt the site.

erronis

(23,471 posts)
2. I'd almost be tempted to say that some people/groups wouldn't mind undermining Wikipedia
Wed Mar 4, 2026, 12:29 PM
4 hrs ago

Free knowledge has been contrary to those in power and to those that want to control it. Think of the religions over the centuries that restricted access to knowledge to the privileged few.

I am so glad to have Wikipedia available at my fingertips whenever I need it. I keep an offline copy (kiwix) just in case those types try to shut it down. (And I regularly contribute $s and sometimes a few additions/corrections.)

quaint

(4,924 posts)
5. I had no idea Wikipedia was paying to include AI crap.
Wed Mar 4, 2026, 02:01 PM
3 hrs ago

I do not donate to many, now one less.

quaint

(4,924 posts)
12. Apologies, I merely posted in the wrong place. The OP is about AI, which is what I meant to reply to.
Wed Mar 4, 2026, 05:06 PM
21 min ago

I am so very sorry I offended you. You offended me by assuming it was other than a mistake.

quaint

(4,924 posts)
11. Apologies, I merely posted in the wrong place. The OP is about AI, which is what I meant to reply to.
Wed Mar 4, 2026, 05:04 PM
23 min ago

I am very sorry to have offended you.

Prairie Gates

(7,800 posts)
6. Rampant errors at best, disinformation more likely
Wed Mar 4, 2026, 02:18 PM
3 hrs ago

We need to stop using this idiotic language of "hallucination." They are not people. They are not hallucinating. They are data processing machines generating systematic error or operating systematically to generate disinformation.

Let's stop agreeing to the language of the Silicon Valley AI bros.

erronis

(23,471 posts)
9. I think any term that gets the idea across to the non-technical is helpful.
Wed Mar 4, 2026, 03:55 PM
1 hr ago

How about brain farts?

Prairie Gates

(7,800 posts)
10. They don't have brains or farts: they're data processing machines
Wed Mar 4, 2026, 04:00 PM
1 hr ago

People understand what errors are. The question is why AI propaganda pretends that people don't.

Latest Discussions»Latest Breaking News»AI Translations Are Addin...