‘Very harmful’: website uses AI to generate northern Indigenous news ‘correspondents’ | CBC News

Articles, images and even writers all created using artificial intelligence on the WE NEWS website

When Kerry Benjoe was an up-and-coming Indigenous journalist in the early 2000s, she says she felt the weight of her community to do justice to their stories. 

“You're educating the public, you're educating your colleagues,” said the interim president of the Indigenous Media Association of Canada.

“There have been a lot of inroads made … trying to create a space where we can finally say these are stories told by us, for us, but also willingly shared with the general public to help correct that false narrative that has been created about us.”

That’s an experience WE NEWS correspondents Kaylee Johns, Warren Lafferty and Leah Qammaniq will never have to face. They’ve been generated using artificial intelligence.

WE NEWS lists profiles for Johns, Lafferty and Qammaniq as “AI correspondents.” Their associated profiles say they are “a member of Yukon First Nation,” “of Dene heritage,” and “an Inuk raised in the North,” respectively. 

Historically, she said, Indigenous people were excluded from the media. 

"Suddenly the space [Indigenous journalists] fought so hard to get, the stories they fought so hard to tell are just completely taken away, they're taken out of it, and I think that is very harmful.”

The website’s About Us page says it “relies on artificial intelligence technologies” and uses “the assistance of AI editorial agents, operating under clear editorial principles: fact verification, transparency about sources, independence of the editorial line, and prompt correction of errors.” 

That process mostly seems to be repurposing articles from legitimate news sources, rewording them and then publishing the edited version with AI-generated images. 

Ahmed Al-Rawi is the director of the Disinformation Project at Simon Fraser University. He says while there’s a “transparency” section at the bottom of each article that says it was written using AI, that isn’t enough. 

“A lot of people might read the article … and think that it's real. The other problem is a lack of attribution, or at least clear attribution, to the original authors,” he said. “It's very unfortunate that they depend on the original work of real journalists in order to generate revenue."