When AI Tools Yield Bad Journalism, Who Is Held Accountable?

When AI Tools Yield Bad Journalism, Who Is Held Accountable?

When AI Tools Yield Bad Journalism, Who Is Held Accountable?

https://www.jezebel.com/ai-in-journalism-tools-pitfalls-reporter-fired-ars-technica-benj-edwards-ap-reporters

Publish Date: 2026-03-04 17:02:00

Source Domain: www.jezebel.com

What exactly is a reporter to do, when the media corporation they’re working for encourages or mandates ever-increasing incorporation of artificial intelligence tools in writing, but those same AI tools—in addition to inherently undermining the need for the employee in question—still can’t be trusted to not fabricate information? Who, in other words, is there to be held accountable when content produced by AI turns out to be flawed, incorrect, or full of bullshit? Do members of management fall on their swords? Or is it the writer who is ultimately most likely to be sacrificed when they’re betrayed by the AI tools they’ve been pushed into utilizing?

We saw something along those lines play out this week, although as in most of these types of stories, blame clearly belongs to multiple parties, and the issues are hardly black and white. A reporter at Ars Technica, whose beat was specifically reporting on AI, was fired after it turned out that a piece he had co-authored contained quotes fabricated by the AI tools he was using. Ars Technica has subsequently retracted the original story entirely, publishing an editor’s note, stating that it was “a serious failure of our standards,” but that they believe it to be an “isolated incident.” Notably, there was no mention of any job losses or discipline of the Condé Nast publication’s editorial leadership—merely the reporter who co-authored the piece, who fell on his sword and took full responsibility.

The writer in question is Benj Edwards, who until recently bore the title of Ars Technica‘s “senior AI reporter.” The article was a write-up on a recent, viral incident from February, in which a coder/programmer named Scott Shambaugh detailed in a blog post how an AI program had essentially threatened him in order to get Shambaugh to favor code it had written. In his own mea culpa, Edwards explained how he had written the…

Source