^{By Natalie Samaha,  AIGP AI governance & literacy Founder of NSM Consultancy}

The promise of AI in journalism is straightforward: assisting with the repetitive tasks so journalists are free to do more reporting. The broad consensus among publishers is that original reporting is more valuable in the AI era. What if the real promise is not time saved, but something else that we are starting to see? 

Currently, backend automation is the dominant AI use case in newsrooms: translation, audio transcription, text-to-speech, tag metadata generation, and code assistance. Doing the same journalism while relying on AI for automating individual tasks.

This pursuit of efficiency is not without cost. According to Press Gazette 2,300 newsrooms jobs were cut in the U.S. and U.K. in the first half of 2026. A 34% increase over 2025’s total of 3,434, if this trend continues. These cuts are largely framed as digital restructuring or reorganisation around digital products.

In the Reuters Institute “Journalism, Media, and Technology Trends and Predictions 2026”, the vast majority (97%) of publishers surveyed considered back-end automation work ‘“important”. At the same time, newsgathering cases (82%) are second most important, with faster coding and product development (81%) also gaining traction.

For David Caswell, founder of StoryFlow, the limits of ‘task automation’ are becoming apparent, and savings of time and money are underwhelming. He sees agentic AI moving into newsgathering, investigations, interviewing, fact-checking and more.

Use cases are emerging of AI becoming part of the reporting workflow itself. They reveal a capacity gain that is more valuable than time saved: newsrooms moving beyond efficiency to journalism they couldn’t attempt before.

The Colonist Report, a small Nigerian outlet, used off-the-shelf generative AI models, mainly ChatGPT and Gemini, to parse 3,000 pages of government documents for their investigation into the 2024 flooding in Rivers State. For editor Elfredah Kevin-Alerechi the use of AI should not focus on what her newsroom already could do, such as on-the-ground reporting, but on providing a capacity it did not have. In this case, being able to analyse a much bigger volume of documents. 

The New York Times’ AI Initiatives Team built CheatSheet, a spreadsheet-based interface that allows reporters to drop datasets and run different preset prompts. In an interview with Nieman Lab in February 2026, Zach Seward, the editorial director for AI initiatives said the tool acts as force multiplier for their reporting and helps to double down on an “existing competitive advantage”.

Filipino journalist and media advisor Jaemark Tordecilla created  “COA Beat Assistant” an AI tool using a custom GPT to help journalists find stories in government audit reports. More than time saving, Tordecilla sees the introduction of custom GPTs as allowing newsrooms to develop apps with a focused scope that can assist journalists in their investigative work. 

In the 1980s and 1990s, personal computers, email, and office software were expected to dramatically reduce administrative work. Instead, research found that they increased the volume of communication and raised expectations of responsiveness. In 1987, Economist Robert Solow said in what became known as the Solow productivity paradox: “You can see the computer age everywhere but in the productivity statistics.” 

Productivity gains from computers and office software started to show years later,  after organisations reorganised around the technology. In comparison, most newsrooms are still in the adoption phase. The above examples show what it would look like when workflows start to organise around the new technology.

The promise of AI in journalism may turn out to be less about time saved than what a newsroom can attempt. This story is still unfolding, and it will be interesting to watch what the industry landscape would look like, and more importantly how journalism will be shaped if newsrooms move past the automation phase.

—

The views and opinions expressed in this article are those of the author and do not necessarily reflect the views of the Euro-Mediterranean Network of Women Journalists.