German publisher Axel Springer says journalists could be replaced by AI

Journalists are at risk of being replaced by artificial intelligence systems like ChatGPT, the CEO of German media group Axel Springer has said.

The announcement was made as the publisher sought to boost revenue at German newspapers Bild and Die Welt and transition to becoming a “purely digital media company”. It said job cuts lay ahead, because automation and AI were increasingly making many of the jobs that supported the production of their journalism redundant.

“Artificial intelligence has the potential to make independent journalism better than it ever was – or simply replace it,” CEO Mathias Doepfner said in an internal letter to employees.

AI tools like the popular ChatGPT promise a “revolution” in information, he said, and would soon be better at the “aggregation of information” than human journalists.

“Understanding this change is essential to a publishing house’s future viability,” said Doepfner. “Only those who create the best original content will survive.”

Axel Springer did not specify how many of its staff could be cut, but promised that no cuts would be made to the number of, “reporters, authors, or specialist editors”.

In his letter to staff, Doepfner said media outlets must focus on investigative journalism and original commentary, while divining the “true motives” behind events would remain a job for journalists.

Axel Springer is not the first news publisher to toy with the use of AI in its content creation. In January, BuzzFeed announced it planned to use artificial intelligence to “enhance” its content and online quizzes.

The published of the UK’s Daily Mirror and Daily Express newspapers is also exploring the use of AI, setting up a working group to look at “the potential and limitations of machine-learning such as ChatGPT”, the group’s chief executive told the Financial Times.

Since its launch in November last year, ChatGPT has amassed more than 100 million users and accelerated a long-predicted reckoning over whether some jobs could be made redundant from artificial intelligence.

The programme can generate highly sophisticated texts from simple user prompts, producing anything from essays and job applications, to poems and works of fiction. ChatGPT is a large-language model, trained by uploading billions of words of everyday text from across the web into the system. It then draws on all this material to predict words and sentences in certain sequences.

However the accuracy of its responses has been called into question. Australian academic have found examples of the system fabricating references from websites and referencing fake quotes.

The use of AI in journalism has proved controversial as well.

Tech website CNET has reportedly been using an AI tool to generate articles that are later scanned by human editors for accuracy before publication. The website acknowledged in January that the program had some limitations, after a report from tech news site Futurism revealed more than half of the stories generated through AI tools had to be edited for errors.

In one example, CNET was forced to issue major corrections to an explainer article on compound interest that contained a number of simple errors.

Reuters contributed to this article