The image reeks of generative artificial intelligence. The warped, watercolor-esque style. Reuters spelled “Redutrs” and El País with an accent over the wrong letter.
It’s currently the main image of the Italian newspaper Il Foglio’s article about AI use in newsrooms — which was written with AI. The image was created with ChatGPT, according to the cutline, as were several others in a new insert in its daily editions called Foglio AI.
“Foglio AI is the first daily newspaper in the world, a real daily newspaper, made every day, the result of discussions, the result of provocations, the result of news, made entirely using artificial intelligence,” the newspaper said in an announcement. “For everything. For the writing, the headlines, the captions, the quotes, the summaries. And sometimes even for the irony.”
I am usually a fan of bold experiments like this, and bristle at the hand-wringing that often comes in response to emerging technology.
But, so far, it’s a case study of how bad AI is at writing the news. Along with the misspellings in images, an article about Donald Trump’s falsehoods has inaccuracies and lacks attribution for anything, two missteps caught by Alexios Mantzarlis, director of the Security, Trust & Safety Initiative at Cornell Tech and former director of the International Fact-Checking Network.
The writing isn’t particularly engaging, and stories end with generic sign-offs. “In conclusion, Donald Trump’s recent statements highlight a tendency to spread inaccurate or misleading information on a wide range of topics. It is crucial for the public and the media to maintain a critical and fact-based approach when evaluating such claims, in order to ensure an informed and transparent public debate.”
As I’ve said before, these tools, if not used properly, write like a hungover college student rushing through a midterm exam.
“(I) don’t have more time to spend on this stunt just now but I expect it to deliver some preposterous results over the month of its pilot,” Mantzarlis wrote in a Bluesky post. “I wish the methodology behind it had been a little clearer.”
I can’t find an AI ethics policy anywhere on Il Foglio’s website. And the article announcing the experiment is vague about the production process, only providing what sounds like a catchy slogan: “We journalists will limit ourselves to asking the questions, in Foglio AI we will read all the answers.”
There are no details about whether humans edited the output. While the outlet discloses its use of ChatGPT and Grok under some images, it doesn’t explain which AI tool was used to write the articles.
To be fair, Il Foglio does ask for audience feedback. And, who knows, their readers may appreciate having more stories to read each week. Writing and reporting be damned.
The writer used Google Translate, Claude and ChatGPT to translate Il Foglio’s articles. Two Italian speakers, including Mantzarlis, were consulted to confirm the context of statements from the newspaper’s announcement.
This piece originally appeared in The Poynter Report, our daily newsletter for everyone who cares about the media. Subscribe to The Poynter Report here.
Comments