As newsrooms face mounting pressure to cut costs, a cautionary tale reveals how careless artificial intelligence use can result in inaccurate journalism.
A special insert titled Heat Index: Your Guide to the Best of Summer, created by King Features (a Hearst syndicate), and published in the Chicago Sun-Times, the Philadelphia Inquirer and other newspapers, went viral for all the wrong reasons.
Readers were quick to spot glaring errors in the section, from invented sources and quotes to attributing books to the wrong authors. It was then revealed that the content was largely generated by AI.
The journalist in charge of the feature admitted to using ChatGPT to draft much of the content, relying on it without adequately verifying sources or quotes. Meanwhile, the newspapers admitted the piece made its way into print with little editorial oversight.
Clarifying the incident, the Chicago Sun-Times drew attention to the rapid transformation journalism and technology are going through, calling this case “a learning moment” for all journalism organisations.
“Our work is valued – and valuable – because of the humanity behind it,” stated the newspaper’s response.
The other side
The mishap comes in stark contrast to the experience reported by Italian daily Il Foglio, which hired ChatGPT to write the world’s first-ever AI-generated newspaper, in a month-long journalistic trial that made headlines.
It was largely deemed a success, while the publishing of the experimental daily was overseen by journalists, who were in charge of putting questions into the AI tool and supervising the answers.
The accomplishment lay within in the human touch; editors refining prompts, correcting biases, and ensuring editorial coherence, proving the irreplaceable role of human journalists in this field.
Il Foglio editor Claudio Cerasa says the chatbot should be treated like a colleague, explaining that the newspaper plans to incorporate AI-written content in a weekly section, with the AI-generated articles clearly labelled as such.
But he preaches against the technology being used to cut corners and replace human intelligence, underlining how the Heat Index debacle reflects the fragility of local journalism: shrinking staff, low pay, and reliance on freelancers under tight deadlines.
“AI is meant to integrate, not replace. Anyone who thinks it is a way to save money is getting it wrong,” Cerasa argues.
AI is not a lifeline
While generative AI promises a way to ease workloads, in practice it often results in unreliable and sloppy material. Incidents like the one with the Heat Index show how ill-used AI can undermine credibility and erode public trust in the media.
AI-powered tools are not the means to restore vital reporting capabilities or serve as a potential lifeline for struggling local newsrooms. Without careful strategy and ethical grounding, opportunities could be squandered – or even be harmful.
Journalism professor Tom Rosenstiel stresses that newsrooms must learn from the internet era’s mistakes: not simply accept AI-generated content, but focus on using AI to build smarter reporting systems and serve community-specific needs.
That means deploying tools that assist journalists, such as bots that alert on local crime trends, automated summaries of public records, and AI-enhanced interviewing platforms, rather than replacing journalists altogether.
His core argument is that success depends on two factors: newsrooms’ willingness to thoughtfully and responsibly adapt and use AI, and their commitment to transparency and editorial standards.
If they do so, AI could free up reporters to dig deeper into important local stories. If they don’t, AI may produce generic, unreliable content that undermines reader trust at a time when trustworthy information is more important than ever.
AI offers local news a “second chance” – but whether the industry is ready to seize it and learn from past mistakes remains to be seen.
[Edited By Brian Maguire | Euractiv’s Advocacy Lab ]
