For decades, the article was the basic unit of news consumption. Audiences opened a news website, read the headline and then worked their way through the full text.
In recent years, however, this model has begun to lose its effectiveness, as artificial intelligence (AI) gradually reshapes how news is accessed and consumed.
Nearly four years after the launch of ChatGPT, in 2026, media experts and newsroom leaders around the world are trying to understand what new rules will shape journalism in the near future.
One of the clearest pictures of what lies ahead comes from a Reuters Institute survey of experts from leading media organisations, including the BBC, The New York Times and The Wall Street Journal. Their responses point to five major directions in which AI is set to transform the news.
1. News will be asked for, not read
The spread of generative AI will continue to fundamentally change how audiences consume news.
Instead of reading articles, people will increasingly follow the news by asking questions to chatbots and AI assistants built into their devices. Rather than clicking on headlines, users will turn directly to AI, asking specific questions that matter to them.
This shift means that news organisations will gradually – and in some cases already – lose control over how their journalism is presented and consumed. There will no longer be a front page, no fixed structure, and no guarantee that an article will be consumed as a complete story. AI systems will break stories apart, extract relevant fragments and reassemble them in response to individual queries.
This is what experts describe as the emergence of an “answer economy”. In this model, audiences will no longer begin by consuming the article itself. Instead, they will seek personalised answers to questions such as: What does this mean for me? How could this affect my work or quality of life? Can this source be trusted?
As a result, journalism will no longer be perceived as a finished product, but increasingly as an informational layer operating inside AI-driven systems.
2. Trust as a prerequisite for survival
Under these new conditions, trust will become the media’s most basic requirement for survival.
Today, almost anyone can generate fake images, videos or audio recordings using AI, and “seeing” no longer means “believing.” Deepfakes and synthetic content are increasingly blurring the line between what is real and what is fake.
As a result, in 2026 news organisations will be expected not only to report, but to prove. Source disclosure, transparency in fact-checking processes and the authentication of visual material will no longer be competitive advantages – they will be baseline journalistic standards.
Some contributors to the Reuters Institute survey even predict that “breaking news” will gradually give way to a race for “breaking verification”.
This shift will present both risks and opportunities. Verification will remain time-consuming and costly, placing additional pressure on already stretched newsrooms. Yet it is also precisely here that professional journalism will be able to distinguish itself from the noise of social media and unverified information and begin to rebuild public trust.
3. Automated newsrooms and AI agents
Another major transformation will unfold inside newsrooms themselves.
While in recent years AI has been used for individual tasks, such as generating headlines, summarising articles or copy editing, 2026 will mark a shift towards automating entire editorial workflows.
AI agents are already capable of carrying out complex processes independently: collecting data, conducting preliminary fact-checks, comparing sources, and even preparing the groundwork for investigative reporting.
Unsurprisingly, this will raise more concerns, particularly among early-career journalists. The labour market will change, and skills that once guaranteed stability in journalism will no longer offer the same security.
At the same time, most experts stress that AI will not replace journalists, but will redefine their role. Routine and repetitive tasks will increasingly be handled by automated systems, while human journalists will be expected to focus on what AI cannot do: understanding context, making ethical judgements and conveying human experience and emotion.
4. Infrastructure matters more than tools
In 2026, media organisations will fully grasp that AI is not primarily a tools problem, but a systems problem.
Newsrooms will no longer be able to simply bolt new AI tools onto outdated workflows and expect meaningful results. Instead, they will begin to rethink their entire infrastructure, from data collection and storage to distribution and monetisation, while investing in AI training across their organisations.
Large media companies are already moving in this direction, building proprietary AI models and personalised news products. For smaller newsrooms, this shift is existential: without technological transformation, remaining competitive will become increasingly difficult.
5. Data journalism gets a second life
AI will give data journalism new momentum.
It enables reporters to work with thousands of documents, classify and compare information and identify connections across large datasets at a scale that was previously almost impossible.
There will, however, remain an important caveat. Archives alone will not produce scoops. Meaningful journalism will depend on access to fresh data – from government institutions, public records, open data platforms and other external sources. Newsrooms that invest in data engineering and analytical capacity will gain a significant advantage.
As these datasets become accessible to the public, citizens themselves will begin querying AI tools to analyse information in their own personal or local contexts. Journalism’s role will therefore shift once again – from merely gathering and publishing information to acting as a mediator between data and the public.
***
Experts can forecast trends, but reality may unfold in unexpected ways. The media now faces a deeply complex question: how to remain trustworthy and useful in a world where information is always available, but truth is not always easy to establish. 2026 will be the year in which those answers are tested.
