Artificial intelligence is present in Europe’s newsrooms, that’s not a surprise, but an Italian newspaper decided to test its full potential by producing and publishing a version made entirely with AI.
The reason? To rethink what we expect from journalism.
Conservative liberal daily Il Foglio made headlines in March when it said it had published the world’s first-ever AI-generated newspaper, where AI is responsible for everything, from the writing, the headlines, the quotes, the summaries, and “sometimes even the irony”.
In this month-long journalistic trial, journalists are limited to simply putting questions into an AI tool and supervising the answers. The content of the experimental daily, Il Foglio AI, occupies four pages and includes articles as well as editorials, written in a more optimistic tone than its traditional counterpart.
A different kind of intelligence
What the experiment seeks to achieve is at the same time fascinating and risky. It wants to test the grounds of how artificial intelligence can turn from its “gaseous”, theoretical state to a solid, practical form. Through the experiment, the newspaper’s staff will gauge the impact it will have on their daily work.
Articles published so far range from a variety of topics: from domestic politics like the upcoming referendum on the Jobs Act, to social issues such as the decrease of romantic relationships among European youth, and even a case-by-case analysis of Donald Trump’s controversial declarations in the past month.
Curiosity and novelty go one step further with an assessment of the first working week of Foglio AI, written by… Foglio AI, in a humorous and sarcastic note, reveals two parallel truths: artificial intelligence can write well, but that does not make it journalism.
AI in the newsroom
With most research on AI in journalism pointing to its potential to enhance newsroom efficiency, freeing journalists for more in-depth reporting, success depends on managing expectations and securing newsroom leadership support.
Two studies have examined AI use at the Associated Press (AP) and the BBC, showing how newsrooms are gradually adapting to it.
At AP, researchers observed the Local News AI initiative, aiming to help small newsrooms integrate AI tools. Those involved in the initiative used AI for tasks such as transcribing city council meetings, summarising police blotters, and automating weather alerts. However, human oversight was needed.
At the BBC, researchers explored journalists’ understanding of AI, finding that many saw themselves as passive users rather than active participants in shaping AI’s role in news. The study highlighted how collaboration with academic researchers could help news organisations avoid bias in their AI adoption.
Should AI take over?
Il Foglio’s experiment at the height of the debate about the use of AI in journalism amid diminishing trust in news, puts journalists’ work as well as readers’ engagement into perspective. Should we let AI do the work simply because it can, and are readers interested in reading what AI has to say?
Foglio AI’s assessment indicates that the newspaper reads well. At times it is “remarkable”, but it is always “consistent”. The experiment shows that AI can be useful and quite a surprising companion in newsrooms, but it can never become the newsroom itself as it cannot yet replicate “humanity in tension”.
While AI rendered routine tasks such as hand-written reports or phone calls to sources unnecessary, it could not, at any moment, replace “a profession that thrives on moods, neuroses, egocentrism, and even mistakes”. As AI itself puts it: “Journalism is the art of knowing when, why, and against whom to write.”
But as legendary journalist Bob Woodward said: “The central dilemma in journalism is that you don’t know what you don’t know.” With AI, that remains true.
[Edited By Brian Maguire | Euractiv’s Advocacy Lab ]
