Whoopsie
The influential, 150-year-old newspaper The Irish Times just apologized for accidentally publishing an AI-generated hoax article — bylined by an entirely fake AI-generated “journalist,” no less — in its Opinion section.
In a statement published Sunday, editor Ruadhán Mac Cormaic apologized for the incident, which he described the incident as a “deliberate and coordinated deception.”
“It was a breach of the trust between the Irish Times and its readers, and we are genuinely sorry,” read the statement. “The incident has highlighted a gap in our pre-publication procedures. We need to make them more robust, and we will.”
“It has also underlined one of the challenges raised by generative AI for news organizations,” Mac Cormaic added. “We, like others, will learn and adapt.”
Welp. So long, reality.
Stir the S***
The anonymous person behind the AI-generated article in question, “Irish women’s obsession with fake tan is problematic,” sounds insufferable. They told the Guardian that they were actively attempting to both “give [their] friends a laugh” and “stir the shit” in the ongoing public discourse about identity politics. Exactly what the internet’s information infrastructure needs right now, right?
According to the Guardian, the plan unfortunately worked, at least to a degree. The piece — which discussed cultural appropriation and the ethics of fake tanning through the purported lens of a fictional Latinx immigrant who grew up in Guayaquil, Ecuador — was reportedly the paper’s second-most read article ever published, and sparked discussion online and on the radio.
Warning Flag
Worse, the Irish Times only realized that it’d published an AI hoax when the author themselves took to Twitter and, from an account attributed to the fake name that they’d published the story under, admitted to the plot.
“Some people have called me an alt-right troll,” the unidentified hoaxer, who described themselves as a nonbinary university student — and who also claims to be from Ireland, not Ecuador, although who knows if any of that is true either — told the Guardian, “but I don’t think that I am.”
The unnamed student added that they’d used OpenAI’s ChatGPT to generate roughly 80 percent of the article, while they used OpenAI’s DALL-E 2 text-to-image to create an image of what they believed to be an accurate representation of a “woke” journalist: “female, overweight, blue hair, business casual clothing, smug expression.”
Gross thing to say! Moving on.
The Irish Times certainly has some egg on its face, and we’d probably advise everyone in media to learn from their mistake. AI tools are pretty much everywhere now — and as a result, sorting through what’s real and what isn’t is likely to get increasingly difficult.