AI
What artificial intelligence is changing in newsrooms — and what it isn't
AI systems are taking over routine tasks in newsrooms, from transcription to topic monitoring. Journalistic responsibility remains with humans.
Few industries discuss artificial intelligence as intensely as journalism. Amid apocalyptic scenarios and promises of salvation, a sober look at what is truly changing in newsrooms is worthwhile.
It is undisputed that AI systems are increasingly taking over routine tasks. They transcribe interviews, translate agency material, suggest headline variations, and sift through large volumes of documents for anomalies. This frees up time for newsrooms — ideally for the work that machines cannot do: investigating, contextualizing, and taking responsibility.
At the same time, new risks arise. Language models generate text fluently, but they do not check facts; they can invent sources and repeat errors with a high degree of persuasiveness. Therefore, reputable outlets adhere to clear rules: AI assistance yes, but a human is responsible for publications, and substantial involvement of AI must be disclosed to the audience.
Changes are also occurring for readers. Automatically generated content floods the internet, making it harder to distinguish between editorially responsible journalism and machine-generated text. Labels, imprint obligations, and reliable senders gain importance — not despite, but because of the new technology.
The core task of journalism is thus not changing: recognizing relevance, verifying claims, and holding power accountable. The tools may change; the responsibility remains.